Paediatric Respiratory Reviews 16 (2015) 81–82

Editorial

Childhood asthma in 2015: The Fuss and the Future

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

The only thing about asthma that is predictable is itspersistence as one of the most common chronic paediatricconditions encountered around the world. It is a condition thatis incompletely understood and so lends itself, somewhatcontroversially, to reclassification [1] and revision of treatmentadvice using the same medications that have been available formany years [2]. The pharmaceutical industry has lost interest to alarge degree in asthma, lured by the prospects of the burgeoningmasses of adults with chronic obstructive pulmonary disease[COPD] following the broader uptake of cigarette smoking in thelast 60 years [3]. The potential rivers of gold for pharmaceuticalcompanies beckon as asthma medications are reinvented as the‘‘new’’ treatment of the Asthma COPD overlap syndrome [ACOS][4].

There are many issues in asthma management that touch us allas respiratory clinicians on a daily basis. In this mini-symposium,familiar clinical situations in diagnosing and managing childrenwith asthma are addressed. The four papers in ‘‘Childhood asthmain 2015: The fuss and the future’’ by eminent respiratoryphysicians explore aspects of asthma care that challenge us toreflect upon our own practices.

The mini-symposium begins on a lighter note with BruceRubin’s paper ‘‘Asthma myths, controversies and dogma’’ [5]. HereRubin considers asthma in the light of theories of its atopicaetiology, management options ranging from ‘‘the soup from a fathen’’ to long-acting beta agonists. The potential for phenotypicchange in the type of asthma with age, growth and environmentalexposures is also discussed. In the paper, Rubin reveals a LatinAmerican myth that if you have a child with asthma and acquire aChihuahua dog, the child will be cured because the asthmasymptoms are transferred to the dog! Somewhat surprisingly,there is no equivalent involving a dingo or kangaroo in Australiathat I am aware of either. There are myths that are easily debunkedbut the challenge for all of us is to deconstruct the dogma that issometimes passed off as data.

In making a diagnosis of asthma we naturally rely on acombination of the history, physical findings and investigations toconfirm or refute the diagnosis. Sophisticated testing is often thedomain of research. Validated testing procedures are ideallycommercialized and introduced after agreement upon interpreta-tion of how one can best utilize the test to distinguish normal fromabnormal. In his paper, ‘‘Exhaled nitric oxide and the managementof childhood asthma-yet another promising biomarker ‘has been’or a misunderstood gem?’’, Steve Turner offers us the insights intothe failure of fractional exhaled nitric oxide [FENO] to translatefrom the benchtop to the clinic in monitoring airway inflammation

http://dx.doi.org/10.1016/j.prrv.2015.01.001

1526-0542/� 2015 Published by Elsevier Ltd.

[6]. With approximately 5% of all children in western countries oninhaled steroids for their asthma [7], the potential for such abiomarker to influence practice is enormous. As previouslydebated in Paediatric Respiratory Reviews, there are clearlyadvocates and skeptics of FENO in the assessment and managementof asthma [8,9], yet Turner brings an interesting sense of equipoisefor FENO and its considerable potential with further work.

In the preventative treatment of asthma over the last twentyyears, little has been offered with more fanfare for children thanmontelukast, a leukotriene receptor antagonist non-steroidalchewable tablet promulgated as ‘‘The asthma pill’’; an alternativeto inhaled steroids. Embraced initially around the world, it carved aniche in the treatment of younger children with milder asthma [10]and as a potential alternative to higher doses of steroids or longacting beta-agonists for poorly controlled persistent asthma[11]. However, studies failed to replicate it efficacy with concernsabout it how useful it has proven to be in comparison to inhaledcorticosteroids for all ages with asthma [12] or in younger childrenwith episodic wheeze [13]. In his paper ‘‘Montelukast in paediatricasthma: where are we now and what still needs to be done’’, AndrewBush reviews the evidence for montelukast and challenges its role inpaediatric practice [14]. Of particular interest, Bush offers practicaladvice through his ‘‘Three step protocol’’ for the prophylactictreatment in preschool wheeze in order to justify ongoing treatmentin a clinical application of personalized medicine.

Marielle Pijnenburg and Stan Szefler move the discussion alongto what is on the horizon for people with asthma. In their article‘‘Personalized medicine in children with asthma’’ they discuss theconcept of individual treatment options in what is clearly aheterogeneous disease [15]. Made-to-order treatment for chronicconditions such as asthma has long been promised yet foundchallenging to deliver. The advent of pharmacogenetics, such aswith the Arg16Gly single nucleotide genotype and stability of carein asthmatics has offered early promise in tailoring treatment to anindividual [16,17]. Together with more targeted use of clinicallypractical biomarkers, better use of electronic medical educationand improved patient communication, much potential exists forimproving patient care.

The journey to overcoming the shortfalls in asthma knowledgeand clinical practice dilemmas is challenging. As clinicians wemust accept small steps as progress. In the meantime, we canconsider the use of daily life trials in personalized medicine asemphasized by Pijnenberg and Szefler, and also by Bush withregard to montelukast. As clinicians, we have come to appreciatethat more medication is not always the answer in medicine but theright medication where needed is.

Editorial / Paediatric Respiratory Reviews 16 (2015) 81–8282

References

[1] Henderson J, Granell R, Heron J, Sheriff A, Simpson A, Woodcock A, et al. Associa-tions of wheezing phenotypes in the first six years of life with atopy, lung functionand airway responsiveness in mid-childhood. Thorax 2008;63:974–80.

[2] Boluyt N, Rottier BL, de Jongste JC, Riemsma R, Vrijlandt EJLE, Brand PLP.Assessment of controversial pediatric asthma management options usingGRADE. Pediatrics 2012;130:e658–68.

[3] Anderson DO, Ferris BG. Role of tobacco smoking in the causation of chronicrespiratory disease. N Engl J Med 1962;267:787–94.

[4] Postma DS, Reddel HK, ten Hacken NH, van der Berge M. Asthma and chronicobstructive pulmonary disease: similarities and differences. Clin Chest Med2014;35:143–56.

[5] Rubin BK. Asthma myths, controversies and dogma. Paediatr Respir Rev2015;16:83–7.

[6] Turner S. Exhaled nitric oxide and the management of childhood asthma-yetanother promising biomarker ‘‘has been’’ or a misunderstood gem? PaediatrRespir Rev 2015;16:88–96.

[7] Turner S, Thomas M, von Zeigenweidt J, Price D. Prescribing trends in asthma: alongitudinal observational study. Arch Dis Child 2009;64:476–83.

[8] Bush A, Eber E. The value of FENO measurement in asthma management-themotion for YES- it’s NO- or the wrong end of the Stick! Paediatr Resp Rev2008;9:127–31.

[9] Franklin PJ, Stick SM. The value of FENO measurement in asthma manage-ment-the motion against FENO measurement to help asthma management-reality bites! Paediatr Respir Rev 2008;9:122–6.

[10] Robertson CF, Price D, Henry R, Mellis C, Glasgow N, Fitzgerald D, et al. Short-course montelukast for intermittent asthma in children: a randomized con-trolled trial. Am J Respir Crit Care Med 2007;175:323–9.

[11] Lemanske Jr RF, Mauger DT, Sorkness CA, Jackson DJ, Boehmer SJ, Martinez FD,et al. Step-ip therapy for children receiving inhaled corticosteroids withuncontrolled asthma. N Engl J Med 2010;362:975–85.

[12] Chauhan BF, Ducharme FM. Anti-leukotriene agents compared to inhaledcorticosteroids in the management of recurrent and/or chronic asthma inadults and children. Cochrane Database Syst Rev 2012;5:CD002314.

[13] Nwokoro C, Pandya H, Turner S, Eldridge S, Griffiths CJ, Vulliamy T, et al.Intermittent montelukast in children aged 10 months to 5 years with wheeze(WAIT trial): a multicentre, randomised, placebo-controlled trial. Lancet RespirMed 2014;2:796–803. http://dx.doi.org/10.1016/S2213-2600(14)70186-9.Epub 2014 Sep 8.

[14] Bush A. Montelukast in paediatric asthma: where are we now and what stillneeds to be done? Pediatr Respir Rev 2015;16:97–100.

[15] Pijnenburg MW, Szefler S. Personalized medicine in children with asthma.Paediatr Respir Rev 2015;16:101–7.

[16] Lipworth BJ, Baso K, Donald HP, Tavendale R, Macgregor DF, Ogston SA, et al.Tailored second-line therapy in asthmatic children with the Arg(16) genotype.Clin Sci (Lond) 2013;124:521–8.

[17] Zuurhout MJ, Vijverberg SJ, Raaijmakers JA, Koenderman L, Postma DS,Koppelman GH, et al. Arg16 ADRB2 genotype increases the risk ofasthma exacerbations in children with a reported use of long acting beta2agonists: results of the PACMAN cohort. Pharmacogenomics 2013;14:1965–71.

Dominic A. Fitzgerald*,

Paediatric Respiratory and Sleep Physician, Clinical Professor,

Sydney Medical School, Discipline of Paediatrics and Child Health,

University of Sydney

*Department of Respiratory Medicine, The Children’s Hospital atWestmead, Sydney, New South Wales, Australia, Locked Bag 4001,

Westmead, New South Wales, Australia 2145.Tel.: +61 2 9845 3397; fax: +61 2 9845 3396

E-mail address: dominic.fitzgerald@health.nsw.gov.au

Paediatric Respiratory Reviews 16 (2015) 83–87

Mini-Sympoisum: Childhood asthma: The fuss and the future

Asthma myths, controversies, and dogma

Bruce K. Rubin *

Department of Pediatrics, Virginia Commonwealth University School of Medicine, Children’s Hospital of Richmond at VCU, Virginia, United States

A R T I C L E I N F O

Keywords:

Asthma

Treatment

Education

Urban myths

Medical misconceptions

S U M M A R Y

Although the symptom complex we call asthma has been well described since antiquity, our

understanding of the causes and therapy of asthma has evolved. Even with this evolution in our

understanding, there are persistent myths (widely held but false beliefs) and dogma (entrenched beliefs)

regarding the causes, classification, and therapy of asthma.

It is sobering that some of the knowledge we hold dear today, will become the mythology of tomorrow.

� 2014 Elsevier Ltd. All rights reserved.

EDUCATIONAL AIMS THE READER WILL BE ABLE TO:

� Understand the importance of a detailed history when considering a diagnosis or erroneous diagnosis of asthma in children.� Appreciate that ‘‘exercise induced asthma [EIA]’’ is a term overused when describing children with dyspnea related to

deconditioning, obesity and anxiety.� Recognize that most children with chronic cough and no other symptoms, rarely have asthma as the cause of their cough� Know that although obesity increases the risk of concomitant asthma, that it is also true that obese children and adults with

dyspnea are often misdiagnosed as having asthma.� Share their understanding of the difference between evidence and myth in the management of childhood asthma.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

INTRODUCTION

Asthma is a disease known since antiquity, with one of the firstdescriptions by Aretaeus of Cappadocia [1]. Maimonides suggested‘‘the soup from a fat hen’’ as a useful therapy in his Treatise on

Asthma [2]. Asthma is now generally defined as recurrent and(somewhat) reversible airflow obstruction due airway inflamma-tion, smooth muscle bronchospasm, and mucus hypersecretion[3]. Given the role of inflammation in asthma, and the ground-breaking work by Steve Rennard and his wife, demonstrating thattraditional home-made chicken soup has anti-inflammatoryproperties in vitro [4], Maimonides’ earlier work shows significantinsight. Over the years, there have been many beliefs aboutasthma. Some of these have been proven to be incomplete orincorrect. Some of these myths are more pervasive amongpatients; others exist as controversies among physicians andscientists. The aim of this paper is to look at a few of these myths,

* Jessie Ball DuPont Distinguished Professor and Chair, Department of Pediatrics,

Professor of Biomedical Engineering, Children’s Hospital of Richmond at Virginia

Commonwealth University, 1001 East Marshall St. PO Box 980646

Tel.: +804 828-9602.

E-mail addresses: brubin@vcu.edu, brubin@mcvh-vcu.edu.

http://dx.doi.org/10.1016/j.prrv.2014.09.001

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

mistakes, controversies, and dogma with the realization that someof the things that we currently believe will be fodder for laughterand scorn in a few years.

EPIDEMIOLOGY AND DEFINITIONS

The definition and epidemiology of asthma is an area that hasbeen rife with controversy with the British, Dutch, and othersweighing in over the years [5]. In 1980, Nick Gross famously asked,‘‘What is this thing called love?’’ with regards to defining asthma[6]. The answer from Prof Gross was that we all know what love(asthma) is but are not always willing to accept another’sdefinition. This is very similar to the definition of pornographyby the US Supreme Court Justice Potter Stewart who in1964 suggested that a precise definition was not possible but, ‘‘Iknow it when I see it’’ [7]. Let’s move on from pornography tosomething much more controversial – the definition andepidemiology of asthma.

Asthma is an infectious disease

In the 1950’s and early 1960’s asthma was considered to be due tochronic bacterial infection and it was proposed that antibiotics

B.K. Rubin / Paediatric Respiratory Reviews 16 (2015) 83–8784

would be effective therapy [8]. Although no specific bacterialpathogens were isolated from the airways of patients with asthma,the practice of using antibiotics led to the unexpected discovery thatthe macrolide antibiotics had immunomodulatory properties quiteseparate from antimicrobial properties [9]. A decade ago, it wasbelieved that many patients with severe asthma were chronicallyinfected with Mycoplasma or Chlamydia, and there was a suspicionthat this might be part of the mechanism of action for macrolideantibiotics in asthma therapy [10]. However, trials of macrolidetherapy in adults with poorly controlled asthma have shown clinicalbenefit unrelated to infection with these organisms [11].

There are now data showing that infection does play a role,probably a key role, in asthma. There appears to be an increasedsusceptibility to human rhinovirus in persons with asthma [12]and human rhinovirus infection is the most frequently reportedcause of acute severe exacerbations of this disease [13]. The role ofRSV in asthma predisposition is far more controversial. Aprospective study was conducted in 27 centers to examine theeffect of palivizumab on recurrent wheeze in infants <36 weeksgestation with and without a family history of atopy in infants whodid (N=146) or did not (N=171) receive palivizumab. There was noeffect of palivizumab on recurrent wheezing in the 90 childrenwith a family history of atopy compared to 130 untreated infantswith atopic families suggesting that RSV perhaps predisposes towheezing in an atopy-independent mechanism [14].

Even more interesting is work on the airway microbiome inhealth and disease. The airways below the larynx were oncethought to be sterile in healthy persons, but it is now known thatthere is a resident microbial population in all airways - themicrobiome. Early studies indicate that loss of diversity in thismicrobiome and predominance of pathogens such as Haemaphilus,may characterize the airway microbiome of patients with asthma[15]. It is not clear to what extent these changes are due to thedisease itself or to the therapy (e.g. inhaled corticosteroids).

We may be coming around to asthma being an infectiousdisease after all.

Asthma is an allergic disease

Over half of children with asthma also have allergies, bothallergies and asthma are associated with a TH2 inflammatoryresponse, and allergens are known to trigger acute asthma. Howevermany patients with asthma do not have allergies, including mostwith adult onset asthma [16]. Thus while it is clear that allergy is oneof the co-morbidities associated with asthma and that treatment ofallergy or allergen avoidance may decrease the severity of asthma, itis probably incorrect to state that asthma is an allergic disease.

Asthma is a bronchospastic disease

Airway smooth muscle hypertrophy and bronchospasm is acharacteristic of asthma, however it is inappropriate to refer toasthma as ‘‘reactive airways disease’’ or as ‘‘bronchospasm’’. Thisabhorrent nomenclature has come about for several reasons. As tobronchospasm; bronchodilator medications are readily availableand extremely effective at relieving acute asthma, the evaluationand diagnosis of asthma often hinges on the demonstration ofbronchial hyperresponsiveness to bronchoprovocation challengeor to the reversal of airflow limitation when patients inhale abronchodilator in the pulmonary function laboratory. However, itis possible to have airway hyperresponsiveness to provocationtesting without having signs or symptoms of asthma [17]. So whilethere is bronchospasm, particularly in acute asthma, asthmacannot be and should not be defined as a bronchospastic disease.

As for the rationale for using the term, ‘‘reactive airwaysdisease’’; I am well aware that there is no cure for ‘‘stupid’’ [18].

Children outgrow asthma

This is a belief sometimes held by general practitioners and bymany parents. As airways grow, and particularly as boys approachadolescence and airway diameter increases, asthma symptoms canabate [19]. However patients with moderate or severe asthmararely have resolution of their symptoms even with goodadherence to controller therapy [20]. It is better that we thinkof asthma as a lifelong illness that can be well controlled, than asmerely a phase of childhood. Children requiring daily therapy forasthma certainly outgrow their paediatrician, but probably nottheir asthma.

There is an asthma epidemic

By reading the newspaper (does anyone still read newspapers?)or the Internet, it is easy to get the impression that there areepidemics of many diseases these days; gluten enteropathy,attention deficit hyperactivity disorder, and asthma to name but afew. In a subspecialty practice of pediatric pulmonary, one is just aslikely to encounter the over diagnosis of asthma as the underdiagnosis [21]. I believe that this is because patients with trueasthma who are adherent to their controller therapy usually dowell and so are rarely referred to a subspecialty practice. It is thosewho are not taking their medication appropriately or do not haveasthma at all that we see in our practice. It is likely that asthma isboth under and over diagnosed but there are no clear epidemio-logic data supporting an ‘‘epidemic’’. In many countries, death fromasthma has been declining over the past two decades due in part, tobetter recognition and better therapy [22].

Studies show that about 30% of adults with physician-diagnosed asthma have been misdiagnosed and that incorrectdiagnosis is more common in children and the elderly [23]. Asthmamisdiagnosis is also more common in obese adults, especially men,who use the resources of emergency departments frequently fortreatment of dyspnea [24]. At the other end of the spectrum,asthma is quite uncommon before the age of 2 years [25] and yetmany children with intermittent viral induced wheezing aretreated with corticosteroids and incorrectly carry a diagnosis ofasthma (or reactive airways disease).

Chronic cough usually means asthma

Many children with asthma also cough, and although there is aphenotype of asthma, referred to as cough variant asthma or coughdominant asthma (my preferred term), this seems to be associatedwith airway mucus hypersecretion and poor response to betaagonist bronchodilators [26]. However, chronic cough as a solesymptom of childhood asthma is rare, representing less than 4% ofthose children with cough that persists for more than 3 weeks [27].

Dyspnea with exercise is usually due to asthma

Poorly controlled asthma can lead to shortness of breath withexercise. However, it has been my observation that the diagnosis of‘‘exercise induced asthma’’ (EIA) is more commonly made in highperforming athletes who have dyspnea when pushing to excel andthat these young athletes almost never have true asthma whenexercise testing is performed appropriately. These patients areoften placed on a short-acting bronchodilator before exercise oreven inhaled corticosteroids (ICS) without appropriate testing.

Seear and colleagues studied 52 children referred for investi-gation of poorly controlled EIA but only 8 (15.4%) fulfilled thediagnostic criteria for EIA defined as a 10% or greater decrease inFEV1 with exercise. 23.1% were unfit, 26.9% had vocal corddysfunction, 13.5% had a habit cough, and 21.1% had no

B.K. Rubin / Paediatric Respiratory Reviews 16 (2015) 83–87 85

abnormalities on clinical or laboratory testing [28]. The initialreported symptoms of wheeze or cough often changed following acareful history, particularly among the elite athletes. These resultsare consistent with studies performed by Weinberger andcolleagues demonstrating that when children or adolescents withdyspnoea on exercise have cardiopulmonary exercise testing, mosthave physiologic limitation of exercise, with very few having trueEIA [29].

This myth has also been associated with another; that childrenwith asthma should not exercise. This is rubbish. One of the maingoals of asthma therapy is to control symptoms well enough thatthe patient can exercise without difficulty related to asthma.Asthma symptoms are bad. Exercise is good.

ALLERGY AND ASTHMA

Allergy immunotherapy is effective for treating asthma

Allergy immunotherapy, given by injection or by sublingualadministration, appears to be effective at reducing the allergicresponse to specific allergens and immunotherapy is beneficial inmany patients for the treatment of classic hay fever. The dataregarding the benefits of immunotherapy for the treatment ofasthma is less robust with low grade evidence of effectiveness inreducing symptoms or preventing disease [30]. However multi-faceted interventions to reduce inhalant and food allergenexposure appears to decrease asthma diagnoses in susceptibletoddlers and young children [31]. Therefore, it is prudent to explainthat immunotherapy can be effective for patients with inade-quately controlled allergy, but is far less effective for the treatmentof asthma. It is likely that regular therapy for persistent asthma willstill be required despite immunotherapy.

Moving to a dry climate will improve asthma

Patients with asthma are sometimes told that if they moved to ahot and dry climate, they would have fewer symptoms. Paradoxi-cally others are told that moving to a coastal city will reducesymptoms. While it is true that in very dry climates there are fewermolds, and very hot climates, such as the desert tend to have fewertrees that produce pollen, often these areas come with their ownset of allergens. The paradox is that some urban centers in thedesert, such as Phoenix in the United States, have become lush withgreenery because the people moving to these climates from morehumid environments miss having green lawn and flowering plantsaround and irrigate heavily to support this plant life.

People with asthma should not have pets. Corollary: Certain breeds of

dogs are hypoallergenic.

It is possible to develop allergies to foreign proteins in thedander (hair and skin) of almost any hairy or furry pet. Provocativedata, consistent with the ‘‘hygiene hypothesis’’, suggest thathaving a dog (but not a cat) in the home at the time a child is bornwill reduce the risk of asthma and allergy in later life [32], whileacquisition of a dog after a year of age does not have the sameprotective effect. This is probably due to changes in the gutmicrobiome associated with early childhood exposure to potentialallergens [33].

It has also been said that certain breeds of dog are less allergenicthan others. A persistent myth in Latin America is that if you have achild with asthma and acquire a Chihuahua dog, the child will becured because the asthma symptoms are transferred to the dog.While there are fur dogs, such as poodles that do not shed, and hairdogs that shed their coat, it does not appear that any breed of dog isless ‘‘allergenic’’ than any other [34].

DRUGS AND THERAPY

The long-term use of ICS is disease modifying

It is an attractive hypothesis that ICS will block asthmaticairway inflammation and thus can decrease the long-term severityof asthma or airway remodeling. Although ICS are the mosteffective medications to treat asthma, there are no data that long-term use is disease modifying.

The early use of inhaled fluticasone propionate for wheezing inpreschool children has no effect on the natural history of asthma orwheeze later in childhood, and does not prevent lung functiondecline or reduce airway reactivity [35]. In a large study ofpreschool children at high risk for asthma, two years of ICS therapydid not change the development of asthma symptoms or lungfunction during a third, treatment-free year [36]. The early use ofICS may also carry some risk of impairing lung growth anddevelopment. Infant rhesus monkeys with ‘‘asthma’’ who weregiven ICS had disruption of postnatal ling growth and differentia-tion of airways and lung parenchyma [37]. Thus it is safe to say thatuse of ICS can decrease the symptoms of asthma but are not diseasemodifying.

ICS are addictive

This is not a myth that is held by many physicians, but somepatients and families are concerned that the long term use of ICSwill decrease the response to medication or make the body‘‘dependent’’ on receiving ICS. These misconceptions may comefrom accurate reports in the lay press that unrestricted antibioticuse can lead to the development of drug resistance in bacteria;while addiction to medications such as barbiturates and opioidsare also commonly knowledge. It is important to recognizethat our patients may not recognize that asthma therapy isdifferent and thus, part of asthma education is anticipating thismisunderstanding.

Medications delivered by jet nebulizer work better and more quickly

than those given by pressurized metered dose inhaler (pMDI) to treat

acute or severe asthma

I have written extensively about this myth [38–40]. When usedappropriately, there is no difference between the effectiveness of abronchodilator medication administered by nebulization whencompared to the same medication given by pMDI [41]. Thedifference is that patients receive much larger doses by nebuliza-tion, are more likely to have greater systemic effects, andadherence to nebulized medication administration is worse [42].

Levalbuterol (levosalbutamol) is safer and more effective than racemic

salbutamol

Levalbuterol or levosalbutamol is the R-enantiomer of theshort-acting b2-adrenergic receptor agonist, salbutamol. Thedevelopment of the single isomer salbutamol was driven byknowledge of receptor pharmacology as the R-enantiomerproduces bronchodilatation while the S-enantiomer does not. Inhumans, there is more than 100 fold greater binding of thebronchodilator R-enantiomer to the beta receptor with almost nobinding of the ineffective S-enantiomer [43]. Studies confirm thatthe benefits and side effects of levosalbutamol are identical tothose of racemic salbutamol; except that levosalbutamol is usuallymore expensive [44]. The lack of difference in reported side effectsis even true in children with cardiac disease given racemic orlevosalbutamol [45]. In one study of very severe, acute asthma inchildren, racemic salbutamol was superior to levosalbutamol with

B.K. Rubin / Paediatric Respiratory Reviews 16 (2015) 83–8786

respect to changes in FEV1 and asthma score, when each wasadministered as continuous nebulization. There was no significantdifference between the drugs with respect to admission rates orside effects [46].

Beta agonists improve mucus clearance

Beta agonists can increase ciliary beat frequency in vivo butmucus clearance is dependent not only on ciliary beat frequency,but also on power, the integrity of the epithelium, and theproperties of mucus. In some species, beta agonists are mucussecretagogues [47]. Well conducted studies by Anderson andDaviskis have shown that inhaled beta agonists do not increasemucociliary clearance in persons with chronic severe asthma [48].

Salbutamol must be given frequently and in high dosage during an

acute asthma exacerbation

Frequent and high dose short acting beta agonists are amainstay of therapy for acute severe asthma, yet frequent and highdosage also carries risks decreasing beta receptor expression,hypokalemia, hyperglycemia, abnormal cardiac rhythm, tremorand anxiety. Furthermore, high dose continuous beta agonisttherapy confers no benefits over usual intermittent aerosoltherapy in adults [49] or in children [50]. Patients dying of asthmain my hospital rarely die from beta agonist deficiency. I suspectthat physicians give more beta agonist than needed to produceoptimal bronchodilatation and that this can contribute tomorbidity. It will be up to investigators with courage to test thisin a randomized clinical trial.

Asthma phenotypes are stable though out life. Corollary: asthma

severity is intrinsic to the patient and needs only to be categorized at

initial presentation.

This is a carry-over from earlier guidelines that suggested thatasthma severity should be evaluated at first diagnosis and that‘‘label’’ would remain with the patient forever. Since our long termgoal is to decrease severity by optimizing control, it doesn’t makesense that severity should be classified only at diagnosis.Therefore, asthma severity scoring is now based on symptomsafter achieving maximal control with medication [3]. In a similarfashion, there is a passion for phenotyping or endotyping asthma,but it is likely that these phenotypes are non-static and can changewith time and growth, co-morbidities, therapy, and environmentalexposures.

CONCLUSION

Although this concludes our brief (and undoubtedly, partial)exploration into myths and dogma related to childhood asthma, itdoes not end our opportunities either to deflate these myths or,unfortunately, our vast abilities to create new myths and dogma.Albert Einstein is quoted as saying, ‘‘Only two things are infinite,the universe and human stupidity, and I’m not sure about theformer’’ [51].

References

[1] Karamanou M, Androutsos G. Aretaeus of Cappadocia and the first clinicaldescription of asthma. Am J Resp Crit Care Med 2011;184:1420–1.

[2] Rosner F. Moses Maimonides’ treatise on asthma. Thorax 1981;36:245–51.[3] Expert Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of

Asthma-Summary Report 2007. J Allergy Clin Immunol. 2007; 120(5 Suppl):S94–138.

[4] Rennard BO, Ertl RF, Gossman GL, Robbins RA, Rennard SI. Chicken soupinhibits neutrophil chemotaxis in vitro. Chest 2000;118:1150–7.

[5] Smolonska J, Koppelman GH, Wijmenga C, et al. Common genes underlyingasthma and COPD? Genome-wide analysis on the Dutch hypothesis. Eur RespirJ 2014. PMID: 24993907.

[6] Gross NJ. What is this thing called love? - or, defining asthma. Am Rev RespirDis 1980;121:203–4.

[7] Justice Potter Stewart (concurring opinion) Jacobellis v. Ohio, 378 U.S. 184(1964).

[8] Finke W. Long-term treatment with antimicrobials and anti-inflammatorysteroids in chronic bronchitis and infectious asthma: Its potential value forrehabilitation and prevention. Am Rev Respir Dis 1960;82:405–8.

[9] Kanoh S, Rubin BK. Macrolides as immunomodulatory medications. Mecha-nisms of action and clinical applications. Clinical Microbiol Rev 2010;23:590–615.

[10] Wong EH, Porter JD, Edwards MR, Johnston SL. The role of macrolides inasthma: current evidence and future directions. Lancet Respir Med 2014;2:657–70.

[11] Sutherland ER, King TS, Icitovic N, et al. A trial of clarithromycin for thetreatment of suboptimally controlled asthma. J Allergy Clin Immunol 2010;126:747–53.

[12] Wark PAB, Johnston SL, Bucchieri F, Powell R, Puddicombe S, Davies DE, Laza-Stanca V, Stephen T, Holgate ST. Asthmatic bronchial epithelial cells have adeficient innate immune response to infection with rhinovirus. J Ex Med2005;201:937–47.

[13] Friedlander SL, Busse WW. The role of rhinovirus in asthma exacerbations. JAllergy Clin Immunol 2005;116:267–73.

[14] Simoes EA, Carbonell-Estrany X, Rieger CH, et al. The effect of respiratorysyncytial virus on subsequent recurrent wheezing in atopic and nonatopicchildren. J Allergy Clin Immunol 2010;126:256–62.

[15] Ege M, von Mutius E. Microbial airway colonization: a cause of asthma andpneumonia? Am J Respir Crit Care Med 2013;188:1188–9.

[16] de Nijs SB, Venekamp LN, Bel EH. Adult-onset asthma: is it really different? EurRespir Rev 2013;22:44–52.

[17] Badier M, Guillot C, Delpierre S. Increased asymptomatic airway hyper-re-sponsiveness in obese individuals. J Asthma 2013;50:573–8.

[18] Fahy JV, O’Byrne PM. ‘‘Reactive airways disease’’. A lazy term of uncertainmeaning that should be abandoned. Am J Respir Crit Care Med 2001;163:822–3.

[19] Bitsko MJ, Everhart RS, Rubin BK. The adolescent with asthma. Paediatr RespirRev 2014;15:146–53.

[20] Tai A, Tran H, Roberts M, Clarke N, Gibson AM, Vidmar S, Wilson J, RobertsonCF. Outcomes of childhood asthma to the age of 50 years. J Allergy Clin Immunol2014;133:1572–8.

[21] Seear M1, Wensley D. Chronic cough and wheeze in children: do they all haveasthma? Eur Respir J 1997;10:342–5.

[22] Akinbami LJ, Moorman JE, Bailey C, Zahran HS, King M, Johnson CA, Liu X.Trends in asthma prevalence, health care use, and mortality in the UnitedStates, 2001-2010. NCHS Data Brief 2012;94:1–8.

[23] Luks VP, Vandemheen KL, Aaron SD. Confirmation of asthma in an era ofoverdiagnosis. Eur Resp J 2010;36:255–60.

[24] Pakhale S, Doucette S, Vandemheen K, et al. A comparison of obese andnonobese people with asthma: Exploring an asthma-obesity interaction. Chest2010;137:1316–23.

[25] Stein RT, Martinez FD. Asthma phenotypes in childhood: lessons from anepidemiological approach. Paediatr Respir Rev 2004;5:155–61.

[26] Schmidt JH, Priftis K, Henke MO, Rubin BK. Secretory hyperresponsiveness andmucus hypersecretion. Chest 2014;146:496–507.

[27] Marchant JM, Masters IB, Taylor SM, Cox NC, Seymour GJ, Chang AB. Evaluationand outcome of young children with chronic cough. Chest 2006;129:1132–41.

[28] Seear MI, Wensley D, West N. How accurate is the diagnosis of exerciseinduced asthma among Vancouver schoolchildren? Arch Dis Child 2005;90:898–902.

[29] Weinberger MM. Etiology of exercise-induced dyspnea: not just exercise-in-duced asthma or vocal cord dysfunction. J Allergy Clin Immunol 2008;121:269.

[30] Erekosima N, Suarez-Cuervo C, Ramanathan M, Kim JM, Chelladurai Y, Segal JB,Lin SY. Effectiveness of subcutaneous immunotherapy for allergic rhinocon-junctivitis and asthma: a systematic review. Laryngoscope 2014;124:616–27.

[31] Maas T, Kaper J, Sheikh A, Knottnerus JA, Wesseling G, Dompeling E, Muris JW,van Schayck CP. Mono and multifaceted inhalant and/or food allergen reduc-tion interventions for preventing asthma in children at high risk of developingasthma. Cochrane Database Syst Rev 2009;(3):CD006480. http://dx.doi.org/10.1002/14651858.

[32] Ownby DR, Johnson CC, Peterson EL. Exposure to dogs and cats in the first yearof life and risk of allergic sensitization at 6 to 7 years of age. J Am Med Assoc2002;288:963–72.

[33] Fujimura KE, Demoor T, Rauch M, Faruqi AA, Jang S, Johnson CC, Boushey HA,Zoratti E, Ownby D, Lukacs NW, Lynch SV. House dust exposure mediates gutmicrobiome Lactobacillus enrichment and airway immune defense againstallergens and virus infection. Proc Nat Acad Sci 2014;111:805–10.

[34] Lockey RF. The myth of hypoallergenic dogs (and cats). J Allergy Clin Immunol2012;130:910–1.

[35] Murray CS, Woodcock A, Langley SJ, Morris J, Custovic A, IFWIN study team.Secondary prevention of asthma by the use of Inhaled Fluticasone propionatein Wheezy INfants (IFWIN): double-blind, randomised, controlled study.Lancet 2006 26;368:754–62.

[36] Guilbert TW, Morgan WJ, Zeiger RS, et al. Long-term inhaled corticosteroidsin preschool children at high risk for asthma. N Engl J Med. 2006; 11;354:1985–97.

B.K. Rubin / Paediatric Respiratory Reviews 16 (2015) 83–87 87

[37] Plopper CG, Joad JP, Miller LA, et al. Lung effects of inhaled corticosteroids ina rhesus monkey model of childhood asthma. Clin Exp Allergy 2012;42:1104–18.

[38] Rottier BL, Rubin BK. Asthma medication delivery: Mists and myths. PaediatrRespir Rev 2013;14:112–8.

[39] Rubin BK. The 36th Annual Donald Egan Memorial Lecture. Air and soul:the science and application of aerosol therapy. Respir Care 2010;55:911–21.

[40] Rubin BK, Fink JB. The delivery of inhaled medication to the young child.Pediatr Clinics N America 2003;50:717–31.

[41] Dolovich MB, Ahrens RC, Hess DR, Anderson P, Dhand R, Rau JL, Smaldone GC,Guyatt G. Device selection and outcomes of aerosol therapy: evidence-basedguidelines. Chest 2005;127:335–71.

[42] Rau JL. Determinants of patient adherence to an aerosol regimen. Respir Care2005;50:1346–56.

[43] Penn RB, Frielle T, McCullough JR, Aberg G, Benovic J. Comparison of R-,S-, andRS-albuterol interaction with human b1 -and b2 -adrenergic receptors. ClinRev Allergy Immunol 1996;14:37–45.

[44] Lotvall J, Palmqvist M, Arvidsson P, et al. The therapeutic ratio of R-albuterol iscomparable with that of RS-albuterol in asthmatic patients. J Allergy ClinImmunol 2001;108:726–31.

[45] Kelly A, Kennedy A, John BM, Duane B, Lemanowicz J, Little J. A comparison ofheart rate changes associated with levalbuterol and racemic albuterol inpediatric cardiology patients. Ann Pharmacother 2013;47:644–50.

[46] Wilkinson M, Bulloch B, Garcia-Filion P, Keahey L. Efficacy of racemic albuterolversus levalbuterol used as a continuous nebulization for the treatment ofacute asthma exacerbations: a randomized, double-blind, clinical trial. JAsthma 2011;48:188–93.

[47] Webber SE, Widdicombe JG. The actions of methacholine, phenylephrine,salbutamol and histamine on mucus secretion from the ferret in-vitro trachea.Agents Actions 1987;22:82–5.

[48] Daviskas E, Anderson SD, Shaw J, Eberl S, Seale JP, Yang IA, Young IH.Mucociliary clearance in patients with chronic asthma: effects of beta ago-nists. Respirology 2005;10:426–35.

[49] Rodrigo GJ1, Rodrigo C. Continuous vs intermittent beta-agonists in thetreatment of acute adult asthma: a systematic review with meta-analysis.Chest 2002;122:160–5.

[50] Krebs SE1, Flood RG, Peter JR, Gerard JM. Evaluation of a high-dose continuousalbuterol protocol for treatment of pediatric asthma in the emergency de-partment. Pediatr Emerg Care 2013;29:191–6.

[51] Perls FS. Ego, Hunger, and Aggression: a Revision of Freud’s Theory andMethod. London: George Allen and Unwin Ltd; 1947: 111.

Paediatric Respiratory Reviews 16 (2015) 88–96

Mini-Sympoisum: Childhood asthma: The fuss and the future

Exhaled nitric oxide and the management of childhood asthma – yetanother promising biomarker ‘‘has been’’ or a misunderstood gem

Steve Turner *

Child Health, University of Aberdeen, Aberdeen, UK

EDUCATIONAL AIMS

� To summarise the literature from observational studies which support the role of fractional exhaled nitric oxide (FENO) as abiomarker for asthma control.� To summarise the results from clinical trials which have used FENO to guide asthma treatment.� To explore why there was an apparent failure to translate FENO from bench to bedside.� To explore how FENO might be used in the future management of childhood asthma

A R T I C L E I N F O

Keywords:

Asthma

Child

Nitric oxide

Respiratory Symptoms

S U M M A R Y

Childhood asthma is a common chronic condition. Approximately five percent of all children in western

countries are prescribed treatment with inhaled corticosteroids (ICS) to prevent asthma symptoms.

Current guidelines advocate titrating ICS dose to symptoms but this approach is not without problem, e.g.

how to discern asthmatic from non-asthmatic symptoms? And when to reduce ICS dose? This review

describes the strengths and weaknesses of fractional exhaled nitric oxide (FENO) as an objective index for

individualising asthma control in children. Epidemiological and mechanistic evidence suggest that FENO

should be a promising biomarker for eosinophilic airway inflammation (a hall mark for asthma) but

somewhat surprisingly, clinical trials in children have not consistently found benefit from adding FENO to

a symptom-based approach to ICS treatment in children. There are a number of reasons why FENO has

apparently failed to translate from promising biomarker to clinically useful tool, and one reason may be a

lack of understanding of what merits a significant intrasubject change in FENO. This review describes the

rise and apparent fall of FENO as biomarker for asthma and then focuses on more recent evidence which

suggest that FENO may prove to have a role in the management of childhood asthma, and in particular

preventing exacerbations.

� 2014 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

A HISTORICAL BACKDROP TO ASTHMA AND NITRIC OXIDE

The search for an asthma control biomarker

Childhood asthma is a very common condition world wide [1]and approximately five percent of all children in western countriesare prescribed inhaled corticosteroids (ICS) to prevent asthmasymptoms [2]. Asthma remains a challenging condition to

* Child Health, Royal Aberdeen Children’s Hospital, Aberdeen, UK, AB25 2ZG.

Tel.: +44 1224 438470.

E-mail address: s.w.turner@abdn.ac.uk.

http://dx.doi.org/10.1016/j.prrv.2014.07.005

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

diagnose and manage in children (and adults) since there is nodefinition, diagnostic test or biomarker to objectively monitordisease control. Historically, several biomarkers have beenevaluated as potential biomarkers for asthma control includingpeak flow, spirometry, bronchial hyperresponsiveness and eosin-ophil cationic protein but these tests all lack sufficient sensitivityand specificity. This review will focus on the potential for fractionalexhaled nitric oxide (FENO) to be a biomarker for childhood asthma.This review will not explore the potential utility of FENO fordiagnosing asthma which has been reviewed elsewhere [3,4]. A asa simple rule low FENO (<10ppb) can be considered a good screento exclude allergic asthma in children aged � five years andconcentrations of �19ppb might have positive predictive value [4]

Table 1Clinically important questions in asthma management where FENO may give insight

Are these asthmatic symptoms in this child with asthma?

Should treatment be stepped up with inhaled corticosteroids or alternative

medications?

When is it appropriate to step down inhaled corticosteroid treatment?

When is it safe to stop treatment with inhaled corticosteroids?

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–96 89

but the interpretation of higher FENO remains challenging and thisis predominantly due to confounding by atopy which leads toelevated FENO independent of asthma.

There is a pressing need for a biomarker for asthma manage-ment in children [5] due to a number of clinically importantquestions to which there are currently no answers (Table 1).Currently the management of asthma is driven by symptoms andat times can be based on trial and error. One example of clinicaluncertainty is the case of a child with asthma symptoms despitetreatment with inhaled steroids – does the clinician increase ICSdose or add in long acting beta agonist or leukotriene receptorantagonist? Children with asthma also get non-asthmatic respira-tory symptoms [6] so how does the clinician deduce whetherrespiratory symptoms in a child with asthma are asthmatic or not?Third and fourth clinical scenarios are the decision-making behindstepping down or stopping ICS treatment in a child with no asthmasymptoms on ICS treatment? Exhaled NO has the potential to giveinsight into these everyday clinical dilemmas.

Exhaled nitric oxide and asthma control, a brief summary of the

evidence

Until the late 1980s, nitric oxide was thought to be just apollutant generated from burning fossil fuels, but was subsequentlyfound to be important to cellular function in many human organsand in 1992 was voted molecule of the year by Science magazine.Nitric oxide, a simple diatomic molecule, proved to be important incellular communication and was the substance previously known asendothelial derived relaxing factor, a potent vasodilator. Nitric oxideis produced by two enzymes. Constitutive nitric oxide synthase(NOS) constantly produces NO at relatively low concentration andthis activity is thought to be important to health and wellbeing; atlow concentrations NO’s properties in the respiratory system mayinclude antimicrobial, immune regulation and possibly bronchodi-lation. The second enzymatic source of NO is inducible NOS which,on stimulation, can produce higher concentrations of NO comparedto constitutive NOS which are associated with disease [7–9]. In the

Table 2Summary of the literature suggesting that exhaled nitric oxide (FENO) may or may not

Studies suggesting FENO may be a good biomarker for childhood asthma

FENO is elevated in children with asthma [13]

Exhaled nitric oxide is positively correlated with

three hallmarks for asthma, sputum eosinophils

[44,81,82] (r=0.5), FEV1 [44] and bronchial

hyperresponsiveness (BHR) [45,46]

Exhaled nitric oxide is positively correlated with

airway eosinophilia after two weeks treatment

with oral corticosteroids (r=0.5) [10]

Elevated FENO is associated with poor asthma

control (r=0.2) [41–43]

FENO rises after withdrawal of ICS and before

symptoms relapse [18]

Treatment with inhaled corticosteroids reduces

FENO in children with asthma [68].

airways, higher concentrations of NO have no homeostatic role andare thought to be secondary to eosinophil inflammation [10]. Thepresence of gaseous nitric oxide in human exhaled breath was firstreported in 1993 [11] and shortly afterwards was found to beelevated in adults with asthma [12]; this observation was replicatedin children four years later [13]. A flurry of scientific activity relatingexhaled nitric oxide to asthma was published during the early 2000sand this indicated both the potential [14,15] and the limitations [16]of using NO in exhaled breath as a biomarker for asthma (Table 2).

With the epidemiology and cellular/molecular work pointing toFENO being a potential biomarker for asthma control in childrenand a standard methodology agreed, a number of studies exploredwhere FENO might be used in asthma management. One studydemonstrated how rising exhaled nitric oxide concentration (usinga threshold concentration of >22ppb) and rising airway eosino-philia (using % eosinophil count as a continuous variable) wereindependently predictive of failure to step down inhaled cortico-steroids in children with stable asthma [17]. A second studymeasured FENO four weeks after cessation of ICS treatment andfound that concentrations in excess of 49ppb had the bestsensitivity (71%) and specificity (93%) for subsequent asthmarelapse [18]. By 2005 clinical trials were under way where FENO

was applied to asthma management as an adjuvant to the standardsymptom-based approach advocated by consensus guidelines.

A standard methodology for measuring NO in exhaled breath

This was agreed by the American Thoracic and EuropeanRespiratory Societies and published in 1999 [19] and revised in2005 [20]. One of the challenges in measuring NO in exhaled breathis flow dependence, i.e. at higher expiratory flows, concentrationsare reduced and vice versa. The flow dependence of exhaled NOdoes give insight into the origin of elevated NO in an individual(broadly from the proximal or distal airways) by deriving flowindependent parameters. Descriptions of derivation of flowindependent parameters and their potential clinical relevance inchildren are available elsewhere [21,22]. The agreed standard wasto measure the fractional exhaled nitric oxide at 50 ml/s. Using thismethodology, a child without asthma would typically have FENO of8-10 parts per billion (ppb) [23] but concentrations might be up to25 ppb [24]. Not only was there evidence to support the paradigmthat FENO was a biomarker for asthma control from epidemiologi-cal, observational and mechanistic studies, FENO measurementscould be made quickly, with minimal discomfort, good reproduc-ibility [25] and results were available within minutes.

be a good biomarker for childhood asthma.

Studies suggesting FENO may NOT be a good biomarker for

childhood asthma

FENO is elevated in atopic non-asthmatic children [45,79] and

in adolescents whose asthma has remitted [80]

Exhaled NO is not related to FEV1 [45] or BHR [48]

FENO is not correlated with asthma control [47]

FENO does not predict relapse after ICS withdrawal [83]

FENO remains elevated in some individuals despite treatment

with ICS [84,85].

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–9690

EXHALED NO AS A BIOMARKER FOR ASTHMA MANAGEMENT INCHILDREN

Results from clinical trials

At the time of writing there have been at least eight trialspublished which explored the clinical utility of FENO in themanagement of asthma in children [26–33]. These randomisedclinical trials compared standard symptom-based managementagainst standard management plus FENO (rather than symptombased versus FENO based management) and each study usedabsolute FENO values to guide changes in treatment (rather thanrelative or personalised FENO values). The clinical trials wereundertaken by groups working independently and inevitablythere is considerable heterogeneity between designs of the trials(Table 3). The lower age limit for inclusion varied between 5 and 12years, one recruited from the community [26] whilst theremainder recruited from hospital clinics [27–33] and some onlyincluded atopic children with asthma [27,30,32,33]. The absoluteFENO values used as cut offs ranged between 10 and 40ppb, sometrials had only one cut off FENO value [27,29,30,32,33], whilstothers had three or four FENO values to trigger escalation in asthmatreatment [26,28] and one employed different single cut offs for anindividual based on their atopic status [31]. One study alsoincluded FEV1 in the decision making algorithm in addition to FENO

[27]. The primary outcome for the studies, upon which the powercalculations were based, were varied and included ICS dose [28–30] FEV1 [27], exacerbations [28,29,31], severity [29] and beingsymptomatic [32,33]. None of the studies observed improvedasthma control among the FENO arms, four found reducedexacerbations [26,29,31,33], two found improved physiologicalmeasurements (i.e. spirometry [27] and bronchial hyperrespon-siveness [30]), two found increased doses of ICS among thoserandomised to FENO guided treatment [26,27] and one foundreduced asthma severity over the course of the trial [29]. One veryrecent study, published only in abstract form at the time of writing[34] reported symptoms free days in 280 children aged 4-18 yearsrandomised to (i) symptom driven treatment (ii) web-basedmonthly monitoring and (iii) symptoms based treatment plus 4monthly FENO measurement; here symptom free days increasedmarginally the FENO arm. Systematic reviews and meta-analysesusing data from some of these studies have concluded that theevidence does not support the addition of FENO to standardsymptom-based management of asthma for day-to-day control[35–37] but one finds evidence for FENO leading to reducedexacerbations [37]. In contrast, at least one expert group arguesthat FENO has an important role in the management of asthma [38].Between evidence synthesis [35–37] and expert opinion [38], arecent report from the National Institute for Clinical Efficacy in theUK [39] has suggested that ‘‘it could be argued that the availableevidence does point towards some benefit to the technology [FENO

measurement]’’ and cites limitations in the current literature asincluding ‘‘cut off values [which] are highly variable and largelybased on derivation studies’’ and ‘‘unclear step-up/step-downprotocols’’.

Meta-analysis

Although this is not a systematic review, the eight papersidentified in section 2.1 are likely to represent most paperspublished in this area and a meta-analysis was undertaken usingstandard software (Review manager 5.2). The outcomes were (i)risk for an individual requiring at least once course of oral

corticosteroids. Details of individuals requiring �1 course of OCSwere provided by the author of one study [28] and was notavailable for a second [29]. Meta-analysis of seven studies

demonstrated that risk for an individual having an exacerbationrequiring OCS was reduced by treatment guided by FENO plussymptoms versus symptoms alone, odds ratio 0.67 [95% CI 0.51,0.88] (Figure 1). One study [26] contributed almost two thirds ofdata for this analysis and substantially influences the overall resultfrom the meta analysis.

(ii) risk for an individual having any exacerbation (however

defined in the study design). The risk for an individual having �1exacerbation of any type could not be determined in two studies(one reported total number of exacerbations [27] and a second didnot report exacerbations [29]); treatment with FENO plussymptoms was associated with an identical reduction in riskcompared to symptoms only as in (i) above (OR 0.67 [95% CI 0.51,0.88].

(iii) ICS dose at the end of the study. Analysis for ICS dose at end ofstudy was complicated by data being presented as median andinterquartile range whereas the software (widely regarded as thegold standard) requires mean and standard deviation values. Datawere transformed to mean and standard deviation [40] assumingthat 25th and 75th centile values were low and high end of therange; these assumptions can be easily challenged and should beconsidered when interpreting the results from this meta-analysis.Data were not available for three studies of which two [29,30]reported (in the text) no increase in dose and one [26] whichreported higher dose ICS (mean difference 119 microg budesonideequivalent [95% CI 49, 189]) associated with treatment guided byFENO. Among the remaining 5 studies there was an overall meanincrease in ICS dose of 106 microg BUD equivalent [95% CI 75, 138],Figure 2. The magnitude of this association is consistent with theone large study which dominated the meta analysis [26] and FENO

guided treatment seems to be associated with an increased in ICSdose of approximately 100 microg BUD equivalent. In addition tothe assumptions about mean and SD values (which resulted in anapparent dose reduction for the FENO arm of the study by de Jongsteet al., [32] where median values in the two arms were equal at 200microg), there is an additional caveat to these results; the resultsare heterogeneous and when adjusted for (using random effects)the mean increase in ICS is 88 microg BUD equivalent [95% CI -10,86].

WHY MIGHT EXHALED NO NOT BE A USEFUL BIOMARKER?

Exhaled NO is poorly specific for asthma

Elevated NO is a biomarker for eosinophilic inflammationrather than for asthma per se and this indirect relationship withasthma may explain why some studies find FENO is an index ofasthma control scores [41–43], FEV1 [44] and bronchial hyperresponsiveness (BHR) [45,46], but FENO is not universallyassociated with control [47], FEV1 [45] or BHR [48]. There is thepossibility that FENO is a more accurate index of asthma control forsome individuals, eg those with atopy, or for individuals wherethere is discordance between symptoms and FEV1. Eosinophilicinflammation may be asymptomatic and this most likely explainsthe relationship between FENO and atopy and bronchial hyperre-activity in children without asthma [45,46,49,50]. It has beenproposed that FENO is merely an index of atopy, i.e. a skin prick test,since concentrations are positively correlated with the number ofskin tests [45] and age at onset of atopy [51] but this is probablyover simplistic since FENO does change acutely after exposure tooral corticosteroid treatment [52], certain foods [53,54], exercise[55] and pollen [56]. What has been recognised is that factors otherthan asthma may acutely and chronically influence NO productionin children (Table 4, Figure 3). Male gender and increasing heightare consistently associated with modest increase in FENO

concentrations and, although children are not likely to grow by

Table 3Details of the six randomised controlled trials comparing standard symptom-based asthma management against standard management plus exhaled nitric oxide (FENO) in

children with asthma.

Study Population details FENO Cut off(s) used

for high FENO

Study design Primary outcome Secondary outcomes

de Jongste

[32]

Aged 6-18 attending

academic centres or

hospitals. Atopic (by

plasma IgE or skin prick

test). Stable mild-

moderate asthma. 151

randomised.

�20 ppb for 6-10 year olds

�25 ppb for >10 year olds

30 week study,

intervention arm made

daily FENO measurements.

Treatment reviewed each

3 weeks by telephone,

physiological testing 1, 3,

5 months and at end of

study

Symptom free days during

last 3 months of trial; this

improved equally in both

arms of the trial.

No difference between

control and intervention

arm for ICS dose, FEV1,

FENO or exacerbations.

Peirsman

[33]

Age range not stated. Mild

to severe asthma

attending hospital clinics.

Atopic (by plasma IgE or

skin prick testing). 99

randomised

�20 ppb 52 week study. FENO and

symptoms reviewed every

three months

Symptom free days; no

difference between

groups

Exacerbation; reduced in

intervention arm (18/49)

compared to the control

arm (35/50).

Fritsch

[27]

Aged 6-18 years. 52

randomised.Attending

hospital clinic. Skin prick

positive.

�20ppb 6 month duration,

assessed each 6 weeks

FEV1 – no difference Exacerbations, mid

expiratory flows, control.

Mid expiratory flow 11%

higher in FENO group.

Increased ICS doses (200

microg/day) in FENO

group.

Petsky

[31]

Aged >4 years 81 children

invited 63 randomised.

Attending hospital clinic.

�10 ppb for non atopic

children

�12 ppb with one positive

skin test

�20 ppb with more than

one positive skin test

12 month study, monthly

visits for four months and

alternate months

thereafter.

Exacerbation – FENO

associated with reduced

exacerbations (19% versus

47%)

Quality of life and

spirometry did not

significantly differ

between groups

Pijnenberg

[30]

Aged 5-18 years. 108

screened 89 randomised.

Attending hospital clinic.

Atopic asthma treated

with ICS.

�30ppb 12 month study with

assessments each 3

months

ICS dose. No difference

between groups.

FENO group had improved

PD20 (1.3 doubling doses),

lower FENO (geometric

mean difference at end of

study 32% lower) and

trend for fewer

exacerbations (20% versus

39%)

Pike

[28]

Aged 6-17 years. 96

screened, 90 randomised.

Attending hospital clinic

with moderate-severe

asthma.

�15ppb

15.1-24.9ppb

�25 ppb

12 month study, assessed

each 2 months

ICS dose and exacerbation.

No difference between

groups.

Spirometry, no difference

between groups.

Szeffler

[26]

Aged 12-20 years. 780

screened. 546

randomised. Inner city

area where �20%

households below poverty

level.

0-20

20.1-30

30.1-40

>40

46 week duration

assessments each 6-8

weeks

Number of days with

symptoms. No difference

between FENO and control

groups

FENO group had:

Mean increased

fluticasone treatment 119

microg/day.

10% reduction in

proportion requiring OCS

Among obese children 0.6

fewer days with

symptoms. For those with

multiple positive skin

tests (ie >9 out of 14

tested) 0.8 fewer days

with symptoms.

Verini

[29]

Aged 6-17 years. 64

children. Referred to

hospital and admitted.

�12 ppb 12 month study with

assessments at baseline

and after 6 and 12 months

Severity score (mean

reduced significantly from

1.1 to 0.6 and 0.8 after 6

and 12 months only in the

FENO group). Exacerbation

(mean number reduced

from 2.0 to 1.0 and 0.8

only in FENO group),

treatment (unchanged in

FENO group but some

evidence of increased

treatment in control arm).

Spirometry – no

difference

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–96 91

more than a few cm between clinic visits, the association withanthropometric measurements challenges the logic behind havingsingle FENO values to trigger changes in ICS throughout childhood;a teenager will grow by as much as 30 cm during puberty and their

FENO value will rise by approximately 5-10 ppb. As an aside, theassociation between height and increased FENO is an interestingobservation since a measurement of concentration should adjustfor size so this is not simply bigger people producing more NO.

Figure 1. A forest plot comparing the effect on exacerbations requiring oral corticosteroid treatment where maintenance treatment is driven by exhaled nitric oxide (ENO)

and symptoms versus symptoms alone.

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–9692

Dietary exposures have been associated with acute changes in FENO

in children [53,54] but these changes are short-lived and of a smallmagnitude. Nitric oxide is derived from the amino acid L-arginineand ingestion of a dose of L-arginine equivalent to two chickenbreasts is associated with a 5 ppb rise in FENO which lasts one hour[54]. Caffeine induces nitric oxide synthase and ingestion of a largedrink of cola leads to a 9ppb increase in FENO after 30 minutes whichresolves after one hour [53]. Inhaled exposures such as second handtobacco smoke [57,58] and poor outdoor air quality [59] areassociated with increased FENO but it is not known how long thesechanges last for. Respiratory infection with virus temporarily affectsFENO values but the nature of this association is not clear; FENO valuesare reduced in infants with respiratory syncitial virus [60] or rhinitis[61] but increased in adults with experimentally induced rhinovirusinfection, FENO rises by approximately 5ppb [62]. There is little directevidence of the effect of viral infection in children; indirect evidencecomes from observations made during exacerbations, precipitatedby rhinovirus, which are associated with elevated FENO [52,63]. Theapparently inconsistent findings between virus infection andchanging FENO might reflect differences in the host response todifferent virus which may be age related and also the retention of NOwithin secretions. Further evidence of almost continuous but smallfluctuations in FENO is evidenced by the diurnal variability inconcentrations [64]; concentrations are less than 1 ppb higher in themorning compared to the afternoon. In addition to variability overminutes and hours, FENO is elevated in children with asthma duringperiods when grass pollen exposure is present [41,56] and also iselevated during the autumn (when moulds cast spores) for thoseexposed to indoor moulds [43]. Children with hayfever haveelevated FENO [65] and concentrations become particularly elevatedduring the spring when compared to those without hayfever [43]. Inaddition to the factors described in Table 4 and Figure 3, intrasubjectvariability in FENO measurements may also be introduced by theapparatus itself. As with all analytical processes, there is variability

Figure 2. A forest plot comparing the effect on inhaled corticosteroid dose at the time of s

symptoms versus symptoms alone.

in repeated measurements using the same apparatus and thisvariability can be reduced by measuring two or three FENO valuesand reporting the mean value [20] but this requires time and alsocosts money. Further apparatus-dependent variability arises whendifferent methods to derive NO are used; one study found anintrasubject difference of 4ppb between devices made by the samemanufacturer [66]. Intrasubject variability becomes considerablygreater when apparatus from different manufacturers are used [67]where a typical difference might be 8ppb but range between -12 and+28ppb. At present it seems sensible to make repeated measure-ments for a given individual using the same apparatus.

Trials were confounded by poor adherence with inhaled corticosteroid

treatment

Adherence to ICS treatment is crucial to the interpretation ofelevated FENO, as it currently is for standard symptom-basedasthma management. Elevated FENO is associated with poorasthma control [41–43] and poor adherence with ICS treatment[26,68], whereas increasing ICS treatment leads to reduced FENO

[68]. Adherence to treatment is always a challenge to measure inasthma, one paper found that typical FENO concentrations foradolescents with adherence was >50% was 24 ppb and was 31ppbfor those with <50% compliance [26]. A second study of 17 childrenfound that compliance with ICS of between 75 and 100% wasassociated with a relative reduction in FENO of 50-100% whereascompliance below 75% was associated with changes in FENO of lessthat 50% [68]. Observations of heterogeneity in FENO response toICS [69,70] might reflect the presence of individuals with high FENO

but little airway eosinophilia, a phenomenon seen in adults [71]but not described in children, or heterogeneity in adherence to ICStreatment. Although there is most likely to be incompleteadherence to ICS in the clinical trials, asthma outcomes improved

tudy exit where maintenance treatment is driven by exhaled nitric oxide (ENO) and

Table 4Factors which are associated with changes in FENO in children independent of asthma

Factor Approximate magnitude of effect

Height Up to 1ppb rise per cm height gained [24]

Dietary exposures Short lived rise of up to 5-10ppb [53,54]

Allergen exposure Rise of up to 50% during birch pollen season [56]

Exposure to second hand smoke Reduction of 100% (26ppb for exposed children versus 56ppb) [57] or absolute reduction of 10ppb [58]

Exposure to poor outdoor air quality Rise of approximately 1ppb 4 hours after each increase of 10mg/m [3] fine particulate exposure (PM2.5) [59]

Genetic variations Variations in genes coding for NOS2 and NOS3 may lead to differences in FENO in adults of 10% [86] or 10ppb [87]

but no association found for NOS1 variant and FENO in children [88]

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–96 93

in both FENO and standard arms of most trials suggesting thatadherence was generally good.

Wrong study design

The clinical trials which have been completed in children to dateall compared standard symptom-based treatment versus standardtreatment plus FENO and perhaps trials should compare symptom-based treatment versus FENO only treatment. This bold study designhas only been used in one trial of adult patients [72] and found thatFENO guided treatment was associated with reduced ICS doses and anon-significant trend for reduced symptoms compared to symptombased management. The poor correlation between asthma controland FENO reported in some studies [41–43] and the lack ofcorrelation in at least one study [47] does question whether asthmatreatment can be guided only by FENO. On the one hand, FENO andsymptoms measure different outcomes and therefore an algorithmwhich captures both outcomes might be better than either alone. Amore conservative approach might argue that there is a too much ofa leap of faith involved in using FENO to guide treatment, and thesymptom-based approach is patient-centred and therefore symp-toms should predominate as the ultimate trigger for changingasthma treatment.

Insufficient power

Although studies justified their sample size by a powercalculation, descriptions of the power calculations do not include

Figure 3. Summary of the asthma-dependent and independent factors associate

a mean or median FENO value and associated variability.Pragmatically, only two published studies randomised more than100 children [26,32] so it is possible that the remaining studiesmay have been underpowered.

Wrong cut offs used

Although increased FENO is associated with adverse asthmaoutcomes in children, the definition of what is ‘‘increased’’ remainsunclear. Evidence from population studies suggests that concen-trations of >35ppb in children are ‘‘high’’ [38] but the question‘‘what is a significant change in FENO for an individual?’’ remainspoorly understood and has been explored in detail elsewhere [73].One early study suggested that a change of 4 ppb might beclinically significant [74] but, as Table 4 demonstrates, there aremany factors other than asthma which can acutely change FENO byan order of at least 4ppb. Furthermore, a rise of 4ppb might beimportant in a child whose previous FENO was 10ppb but not for asecond individual whose FENO was 20ppb and relative change inFENO seems a more meaningful method for interpreting repeatedmeasurements. Recent studies in adults have suggested that arelative change of <30% is unlikely to be clinically relevant [75] anda change from poor control to good control was associated with aFENO reduction of greater than 35% [76]. Having a ‘‘significant’’magnitude of change in FENO of 30-35% would be consistent with aclinically meaningful change in bronchial hyperreactivity (ahallmark for asthma and correlated with FENO) of half a doublingdose [77]. In children, a FENO rise of 60% from baseline (with 95%

d with increased or reduced concentrations of exhaled nitric oxide (FENO).

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–9694

confidence intervals of approximately 25, 140) was associatedwith an exacerbation [63] and by extrapolation, a rise in FENO ofless than 60% might be indicative of increasing symptoms. Aclinical practical guideline published by the American ThoracicSociety in 2011 [38] acknowledged a weak evidence base andcautiously recommended that a rise in FENO of >20% or (inchildren) >20ppb may be significant and that a minimallyimportant reduction in FENO was >20% for those with a FENO of�50ppb and <10ppb for those for those with lower values. In theadult literature there has been interest in expressing FENO as apercentage of predicted but this option is losing favour, mostly dueto lack of precision and to differences between referencepopulations raising the question of which reference is the bestfor a given population? A fourth method to express FENO is a aspercentage of lowest value and is measured after a two weekcourse of oral corticosteroids, but this has an associated morbidity,might yield a low FENO value which cannot be achieved with ICStreatment and should be reserved for use only in special casesunder expert supervision. Of the four methods described,percentage difference seems best suited for individualisingtreatment since this recognises the relatively wide range of valueswithin a population of children.

Insight into intrasubject variability

One recent study has given insight into the question ‘‘what isa significant change in FENO?’’ [43]. 178 children were recruited,of whom 47 had asthma, in a community-based observationalstudy where FENO was measured over six two-month intervals.The difference between paired FENO measurements wasexpressed as an absolute value and limits of agreement. Asmight be expected, the limits of agreement for paired FENO

measurements were greater for those with higher initialconcentrations. Average FENO values were stable over eightmonths but did become significantly higher over a ten monthinterval, presumably due to the children becoming taller.Asthma was associated with elevated FENO in this population(27ppb versus 10 ppb for non-asthmatic) but when both timeand baseline FENO value were considered, asthma was notindependently associated with change in FENO value. As a roughrule of thumb, the authors suggested that FENO values may riseby up to 100% of the previous measurements over two to fourmonths, independently of asthma. For example, in the 40children with initial FENO between 11 and 20 ppb (median value14ppb) the upper limits of agreement for measurements takenat a two and four month interval were +22ppb and +14 ppbrespectively. As might be expected over time (and regression tothe mean), low initial FENO concentrations became higher whilsthigher concentrations became lower; thus the lower limits ofagreement over two and four months for children whose initialFENO was 21-30 ppb were -19 and -25ppb. In keeping with thesuggestion that a more permissive approach to interpretation ofFENO values, a more liberal algorithm which allowed FENO

concentrations to rise by up to 100% (from 16 to 29ppb) wasfound to be effective in reducing exacerbations and improvingquality of life among pregnant women [78].

In addition to describing variability in FENO over time, this studyrelated FENO to asthma control (both present and future) and alsoto environmental exposures which might affect FENO values [43].There was weak correlation between FENO and current and futureasthma control measured over a four month interval (correlationcoefficient approximately 0.2). Compared with maintained goodasthma control over two months, children who were poorlycontrolled but became well controlled had elevated FENO; incontrast, neither those who had good asthma control whichbecame poorly controlled nor those whose asthma control

remained poor had elevated FENO. These observations suggestedthat elevated FENO is an index of poor current control but not poorcontrol in two month’s time. Additionally the findings suggestedthat the mechanism for persistently poorly controlled symptomsin children with asthma may not involve eosinophilic airwayinflammation.

FUTURE RESEARCH DIRECTIONS - SO WHERE DO WE GOBEYOND 2014 WITH FENO?

It is too early to consign FENO to the dust bin where failedbiomarkers for asthma are placed. There is still sufficientevidence to indicate that FENO may have a role in helping toaddress the current situation where there are too many childrentreated with inappropriately high doses of inhaled corticoste-roids and conversely, too many children with poorly controlledasthma whose quality of life can be improved with ICS treatment.The inconsistency between the epidemiology and mechanisticstudies (supportive of a role for FENO in asthma management)and the clinical trials to date (which are generally not supportiveof adding FENO to standard symptom-based management)suggests either FENO lacks precision or we have not properlyunderstood how to interpret FENO as a clinical tool. Time willshow whether FENO does have role or not in the management ofchildhood asthma. If FENO does prove to have a role in themanagement of childhood asthma then clinicians will have toplace trust in FENO since guidelines will have to use FENO to steptreatment down as well as up. Now that insight is being gainedinto what merits a significant change in FENO, clinical trials areneeded which test cut offs to treatment algorithms. Futureclinical trials designed to use FENO to improve asthma outcomesmight consider the following:

1. Comparing symptom based management and FENO only basedmanagement. This might follow in the success of trialscomparing symptoms versus FENO plus symptoms; the apparentfailure of previous studies will understandably make cliniciansvery cautious in using only FENO to guide treatment.

2. Careful attention to treatment adherence. This needs to beintegral to clinical trials since poor adherence has great potentialto mask any true clinical benefit but in the long term, FENO mayprove to give the clinician insight into adherence.

3. What is the ‘‘best’’ outcome. At present, the evidence wouldsuggest that FENO may have a greater influence in reducingexacerbations rather than improving day-to-day control ofsymptoms. It is possible that one algorithm may lead to bettercontrol and another to fewer exacerbations for a givenindividual. On a practical note, having symptom control as anoutcome and part of the algorithm is a potential flaw in studydesign.

4. Absolute versus relative FENO values. There is sufficient evidenceto categorise individuals as having high FENO on study entry butmore work is required in establishing whether cut offs forsecond and subsequent FENO values should be absolute orpercent of previous values.

5. Algorithms could use FENO to guide treatment step up optionsfor individuals with uncontrolled asthma despite compliancewith ICS treatment, i.e. to further increase ICS or use alternative‘‘add ons’’, as has been applied in adults [78].

6. Algorithms could use FENO to step down ICS treatment, evenwhen (non-asthmatic) symptoms are present.

7. Clinical setting. Childhood asthma is a condition which is mostlymanaged in the community and trial design should ideallyreflect this and aspire to an ideal of easily delivered personalisedtreatment algorithms.

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–96 95

8. Preschool children. Methodologies are required to allow FENO tobe measured in younger children – currently FENO can bemeasured in children aged 5-6 years.

CONFLICTS OF INTEREST

Dr Turner has completed three studies where consumableswere provided by Aerocrine.

References

[1] Lai CK, Beasley R, Crane J, et al. Global variation in the prevalence and severityof asthma symptoms: phase three of the International Study of Asthma andAllergies in Childhood (ISAAC). Thorax 2009;64:476–83.

[2] Turner S, Thomas M, von Ziegenweidt J, Price D. Prescribing trends in asthma: alongitudinal observational study. Arch Dis Child 2009;94:16–22.

[3] Ludviksdottir D, Diamant Z, Alving K, Bjermer L, Malinovschi A. Clinical aspectsof using exhaled NO in asthma diagnosis and management. Clin Respir J2012;6:193–207.

[4] Brodlie M, McKean MC. Exhaled nitric oxide in the diagnosis of childhoodasthma. Brit Med J 2009;339:b5418.

[5] Szefler SJ, Wenzel S, Brown R, et al. Asthma outcomes: biomarkers. J Allergy ClinImmunol 2012;129.

[6] Turner S. An asthmatic child with a troublesome cough. Brit Med J2011;342:c6846.

[7] Guo FH, Comhair SA, Zheng S, et al. Molecular mechanisms of increased nitricoxide (NO) in asthma: evidence for transcriptional and post-translationalregulation of NO synthesis. Journal of Immunology 2000;164:5970–80.

[8] Hamid Q, Springall DR, Riveros-Moreno V, et al. Induction of nitric oxidesynthase in asthma. Lancet 1993;342:1510–3.

[9] Lane C, Knight D, Burgess S, et al. Epithelial inducible nitric oxide synthaseactivity is the major determinant of nitric oxide concentration in exhaledbreath. Thorax 2004;59:757–60.

[10] Payne D, Adcock I, Wilson N, Oates T, Scallan M, Bush A. Relationship betweenexhaled nitric oxide and mucosal eosinophilic inflammation in children withdifficult asthma, after treatment with oral prednisolone. Am J Respir Crit CareMed 2001;164:1376–81.

[11] Borland C, Cox Y, Higenbottam T. Measurement of exhaled nitric oxide in man.Thorax 1993;48:1160–2.

[12] Alving K, Weitzberg E, Lundberg J. Increased amount of nitric oxide in exhaledair of asthmatics. Eur Respir J 1993;6:1368–70.

[13] Nelson BV, Sears S, Woods J, et al. Expired nitric oxide as a marker forchildhood asthma. Journal of Pediatrics 1997;130:423–7.

[14] Kharitonov SA, Barnes PJ. Does exhaled nitric oxide reflect asthma control?Yes, it does! Am J Respir Crit Care Med 2001;164(5):727–8.

[15] de Jongste JC. Yes to NO: the first studies on exhaled nitric oxide-drivenasthma treatment. Eur Respir J 2005;26:379–81.

[16] Franklin PJ, Stick SM. The value of FeNO measurement in asthma manage-ment: the motion against FeNO to help manage childhood asthma–realitybites. Paediatric Respiratory Reviews 2008;9:122–6.

[17] Zacharasiewicz A, Wilson N, Lex C, et al. Clinical Use of Noninvasive Measure-ments of Airway Inflammation in Steroid Reduction in Children. Am J RespirCrit Care Med 2005;171:1077–82.

[18] Pijnenburg MW, Hofhuis W, Hop WC, de Jongste JC. Exhaled nitric oxidepredicts asthma relapse in children with clinical asthma remission. Thorax2005;60:215–8.

[19] American Thoracic, Society. Recommendations for the standardized proceduresfor the online and offline measurement of exhaled nitric oxide and nasal nitricoxide in adults and children. Am J Respir Crit Care Med 1999;160:2104–17.

[20] American Thoracic, Society. Recommendations for Standardized Proceduresfor the Online and Offline Measurement of Exhaled Lower Respiratory NitricOxide and Nasal Nitric Oxide. Am J Respir Crit Care Med 2005;171:912–30.

[21] George SC, Hogman M, Permutt S, Silkoff PE. Modeling pulmonary nitric oxideexchange. J Appl Physiol 2004;96:831–9.

[22] Paraskakis E, Brindicci C, Fleming L, et al. Measurement of bronchial andalveolar nitric oxide production in normal children and children with asthma.Am J Respir Crit Care Med 2006;174:260–7.

[23] Baraldi E, Azzolin NM, Cracco A, Zacchello F. Reference values of exhaled nitricoxide for healthy children 6-15 years old. Pediatric Pulmonology 1999;27(1):54–8.

[24] Buchvald F, Baraldi E, Carraro S, et al. Measurements of exhaled nitric oxide inhealthy subjects age 4 to 17 years. J Allergy Clin Immunol 2005;115:1130–6.

[25] Napier E, Turner SW. Methodological issues related to exhaled nitric oxidemeasurement in children aged four to six years. Pediatr Pulmonol 2005;40:97–104.

[26] Szefler SJ, Mitchell H, Sorkness CA, et al. Management of asthma based onexhaled nitric oxide in addition to guideline-based treatment for inner-cityadolescents and young adults: a randomised controlled trial. Lancet2008;372:1065–72.

[27] Fritsch M, Uxa S,Horak Jr F, et al. Exhaled nitric oxide in the management of childhoodasthma: a prospective 6-months study. Pediatr Pulmonol 2006;41:855–62.

[28] Pike K, Selby A, Price S, et al. Exhaled nitric oxide monitoring does not reducedexacerbation frequency or inhaled corticosteroid dose in paediatric asthma: arandomised controlled trial. Clin Respir J 2013;7:204–13.

[29] Verini M, Consilvio NP, Di Pillo S, et al. FeNO as a Marker of Airway Inflam-mation: The Possible Implications in Childhood Asthma Management. J Allergy2010;691425. http://dx.doi.org/10.1155/2010/691425.

[30] Pijnenburg MW, Bakker EM, Hop WC, de Jongste JC. Titrating steroids onexhaled nitric oxide in children with asthma: a randomized controlled trial.Am J Respir Crit Care Med 2005;172:831–6.

[31] Petsky H, Li AM, Au CT, Kynaston JA, Turner C, Chang AB. Management basedon exhaled nitric oxide levels adjusted for atopy reduces asthma exacerbationsin children: A dual centre randomized controlled trial. Pediatr Pulmonol 2014.http://dx.doi.org/10.1002/ppul.23064.

[32] de Jongste JC, Carraro S, Hop WC, Group CS, Baraldi E. Daily telemonitoring ofexhaled nitric oxide and symptoms in the treatment of childhood asthma. Am JRespir Crit Care Med 2009;179:93–7.

[33] Peirsman EJ, Carvelli TJ, Hage PY, et al. Exhaled Nitric Oxide in ChildhoodAllergic Asthma Management A Randomised Controlled Trial. Pediatr Pulmonol2013. http://dx.doi.org/10.1002/ppul.22873.

[34] Voorend-van Bergen S, Vaessen-Verberne A, Landstar A, et al. FeNO and web-based monitoring in paediatric asthma management; the BATMAN study. EurRespir J 2013;42(Suppl 57):629s.

[35] Petsky HL, Cates CJ, Lasserson TJ, et al. A systematic review and meta-analysis:tailoring asthma treatment on eosinophilic markers (exhaled nitric oxide orsputum eosinophils). Thorax 2010. http://dx.doi.org/10.1136/thx.2010.135574.

[36] Jartti T, Wendelin-Saarenhovi M, Heinonen I, Hartiala J, Vanto T. Childhoodasthma management guided by repeated FeNO measurements: a meta-anal-ysis. Paediatr Respir Rev 2012;13:178–83.

[37] Mahr TA, Malka J, Spahn JD. Inflammometry in pediatric asthma: a review offractional exhaled nitric oxide in clinical practice. Allergy Asthma Proc2013;34:210–9.

[38] Dweik RA, Boggs PB, Erzurum SC, et al. An official ATS clinical practiceguideline: interpretation of exhaled nitric oxide levels (FENO) for clinicalapplications. Am J Respir Crit Care Med 2011;184:602–15.

[39] National Institute for Clinical Excellence. Measuring fractional exhaled nitricoxide concentrations in asthma -NIOX MINO, NIOX VERO and NO breath:-diagnostics consultation document. http://www.nice.org.uk/nicemedia/live/13864/65618/65618.pdf (accessed 23-12-2013) 2013.

[40] Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from themedian, range, and the size of a sample. BMC Med Res Methodol 2005;5(13).

[41] Roberts G, Hurley C, Bush A, Lack G. Longitudinal study of grass pollenexposure, symptoms, and exhaled nitric oxide in childhood seasonal allergicasthma. Thorax 2004;59:752–6.

[42] Piacentini GL, Peroni DG, Bodini A, et al. Childhood Asthma Control Test andairway inflammation evaluation in asthmatic children. Allergy 2009;64:1753–7.

[43] Cutts R, Turner S. Longitudinal measurements of exhaled nitric oxide inchildren - what is a significant change in FENO? Pediatr Allergy Immunol2013;24:540–8.

[44] Pontin J, Blaylock M, Walsh GM, Turner SW. Sputum eosinophil apoptotic rateis positively correlated to exhaled nitric oxide in children. Pediatr Pulmonol2008;43:1130–4.

[45] Franklin PJ, Turner SW, Le Souef PN, Stick SM. Exhaled nitric oxide and asthma:complex interactions between atopy, airway responsiveness and symptoms ina community population of children. Thorax 2003;58:1048–52.

[46] Steerenberg PA, Janssen NA, de Meer G, et al. Relationship between exhaledNO, respiratory symptoms, lung function, bronchial hyperresponsiveness, andblood eosinophilia in school children. Thorax 2003;58(3):242–5.

[47] Waibel V, Ulmer H, Horak E. Assessing asthma control: symptom scores, GINAlevels of asthma control, lung function, and exhaled nitric oxide. PediatrPulmonol 2012;47:113–8.

[48] Silvestri M, Spallarossa D, Battistini E, Brusasco V, Rossi GA. Dissociationbetween exhaled nitric oxide and hyperresponsiveness in children with mildintermittent asthma. Thorax 2000;55:484–8.

[49] van Den Toorn LM, Prins JB, Overbeek SE, Hoogsteden HC, de Jongste JC.Adolescents in clinical remission of atopic asthma have elevated exhaled nitricoxide levels and bronchial hyperresponsiveness. American Journal of Respira-tory & Critical Care Medicine 2000;162:953–7.

[50] Frank TL, Adisesh A, Pickering AC, et al. Relationship between exhaled nitricoxide and childhood asthma. American Journal of Respiratory & Critical CareMedicine 1998;158(4):1032–6.

[51] Turner SW, Heaton T, Rowe J, et al. Early-onset atopy is associated withenhanced lymphocyte cytokine responses in 11-year-old children. Clin ExpAll 2007;37:371–80.

[52] Baraldi E, Azzolin NM, Zanconato S, Dario C, Zacchello F. Corticosteroidsdecrease exhaled nitric oxide in children with acute asthma. Journal ofPediatrics 1997;131(3):381–5.

[53] Abuzayan I, Paraskakis E, Turner SW. Changes to exhaled nitric oxide inasthmatic children after drinking a caffeine-containing cola drink. PediatrPulmonol 2010;45:1228–32.

[54] Abuzayan I, Turner SW. Changes in exhaled nitric oxide after ingestion of L-arginine in children: a pilot study. Pediatr Pulmonol 2010;45:236–40.

[55] Petsky H, Kynaston JA, McElrea M, Turner C, Isles A, Chang AB. Cough andexhaled nitric oxide levels:what happens with exercise? Frontiers Pediatr2013. http://dx.doi.org/10.3389/fped.2013.00030.

[56] Vahlkvist S, Sinding M, Skamstrup K, Bisgaard H. Daily home measurements ofexhaled nitric oxide in asthmatic children during natural birch pollen expo-sure. J Allergy Clin Immunol 2006;117:1272–6.

S. Turner / Paediatric Respiratory Reviews 16 (2015) 88–9696

[57] Laoudi Y, Nikasinovic L, Sahraoui F, Grimfeld A, Momas I, Just J. Passivesmoking is a major determinant of exhaled nitric oxide levels in allergicasthmatic children. Allergy 2010;65:491–7.

[58] Warke TJ, Mairs V, Fitch PS, Ennis M, Shields MD. Possible association betweenpassive smoking and lower exhaled nitric oxide in asthmatic children. ArchEnviron Health 2003;58:613–6.

[59] Mar TF, Jansen K, Shepherd K, Lumley T, Larson TV, Koenig JQ. Exhaled nitricoxide in children with asthma and short-term PM2.5 exposure in Seattle. EnvirHealth Perspect 2005;113:1791–4.

[60] Gadish T, Soferman R, Merimovitch T, Fireman E, Sivan Y. Exhaled nitric oxidein acute respiratory syncytial virus bronchiolitis. Arch Pediatr Adolesc Med2010;164:727–31.

[61] Franklin PJ, Turner SW, Hall GL, Moeller A, Stick SM. Exhaled nitric oxide isreduced in infants with rhinorrhea. Pediatr Pulmonol 2005;39:117–9.

[62] de Gouw HW, Grunberg K, Schot R, Kroes AC, Dick EC, Sterk PJ. Relationshipbetween exhaled nitric oxide and airway hyperresponsiveness followingexperimental rhinovirus infection in asthmatic subjects. Eur Respir J1998;11:126–32.

[63] van der Valk RJ, Baraldi E, Stern G, Frey U, de Jongste JC. Daily exhaled nitricoxide measurements and asthma exacerbations in children. Allergy 2012;67:265–71.

[64] Stark H, Purokivi M, Kiviranta J, Randell J, Tukiainen H. Short-term andseasonal variations of exhaled and nasal NO in healthy subjects. Respir Med2007;101:265–71.

[65] Jouaville LF, Annesi-Maesano I, Nguyen LT, Bocage AS, Bedu M, Caillaud D.Interrelationships among asthma, atopy, rhinitis and exhaled nitric oxide in apopulation-based sample of children. Clin Exp All 2003;33:1506–11.

[66] McGill C, Malik G, Turner SW. Validation of a hand-held exhaled nitric oxideanalyzer for use in children. Pediatr Pulmonol 2006;41:1053–7.

[67] Kapande KM, McConaghy LA, Douglas I, et al. Comparative repeatability of twohandheld fractional exhaled nitric oxide monitors. Pediatr Pulmonol2011;47:546–50.

[68] Beck-Ripp J, Griese M, Arenz S, Koring C, Pasqualoni B, Bufler P. Changes ofexhaled nitric oxide during steroid treatment of childhood asthma. Eur Respir J2002;19:1015–9.

[69] Buchvald F, Eiberg H, Bisgaard H. Heterogeneity of FeNO response to inhaledsteroid in asthmatic children. Clin Exp All 2003;33:1735–40.

[70] Pijnenburg MW, Bakker EM, Lever S, Hop WC, De Jongste JC. High fractionalconcentration of nitric oxide in exhaled air despite steroid treatment inasthmatic children. Clin Exp All 2005;35:920–5.

[71] Shaw DE, Berry MA, Thomas M, et al. The use of exhaled nitric oxide to guideasthma management: a randomized controlled trial. Am J Respir Crit Care Med2007;176:231–7.

[72] Smith AD, Cowan JO, Brassett KP, Herbison GP, Taylor DR. Use of exhaled nitricoxide measurements to guide treatment in chronic asthma. N Engl J Med2005;352:2163–73.

[73] Gibson PG. Using fractional exhaled nitric oxide to guide asthma therapy:design and methodological issues for ASthma TReatment ALgorithm studies.Clin Exp All 2009;39:478–90.

[74] Kharitonov SA, Gonio F, Kelly C, Meah S, Barnes PJ. Reproducibility of exhalednitric oxide measurements in healthy and asthmatic adults and children. EurRespir J 2003;21:433–8.

[75] Michils A, Louis R, Peche R, Baldassarre S, Van Muylem A. Exhaled nitric oxide as amarker of asthma control in smoking patients. Eur Respir J 2009;33:1295–301.

[76] Hewitt RS, Modrich CM, Cowan JO, Herbison GP, Taylor DR. Outcomes usingexhaled nitric oxide measurements as an adjunct to primary care asthmamanagement. Primary Care Respir J 2009;18:320–7.

[77] Weatherall M, Finlgeton J, Eyers S, Beasley R. A half doubling does change inbronchial hyperresponsiveness in a populaiton represents an important dif-ference. Translational Respir Med 2013;1:4.

[78] Powell H, Murphy VE, Taylor DR, et al. Management of asthma in pregnancyguided by measurement of fraction of exhaled nitric oxide: a double-blind,randomised controlled trial. Lancet 2011;378:983–90.

[79] Steerenberg PA, Janssen NA, de Meer G, et al. Relationship between exhaledNO, respiratory symptoms, lung function, bronchial hyperresponsiveness, andblood eosinophilia in school children. Thorax 2003;58:242–5.

[80] van Den Toorn LM, Prins JB, Overbeek SE, Hoogsteden HC, de Jongste JC.Adolescents in clinical remission of atopic asthma have elevated exhaled nitricoxide levels and bronchial hyperresponsiveness. American Journal of Respira-tory & Critical Care Medicine 2000;162(3 Pt 1):953–7.

[81] Piacentini GL, Bodini A, Costella S, et al. Exhaled nitric oxide and sputumeosinophil markers of inflammation in asthmatic children. European Respira-tory Journal 1999;13(6):1386–90.

[82] Warke TJ, Fitch PS, Brown V, et al. Exhaled nitric oxide correlates with airwayeosinophils in childhood asthma. Thorax 2002;57:383–7.

[83] Cabral AL, Vollmer WM, Barbirotto RM, Martins MA. Exhaled nitric oxide as apredictor of exacerbation in children with moderate-to-severe asthma: aprospective, 5-month study. Ann Allergy Asthma Immunol 2009;103:206–11.

[84] Buchvald F, Eiberg H, Bisgaard H. Heterogeneity of FeNO response to inhaledsteroid in asthmatic children. Clin Exp All 2003;33(12):1735–40.

[85] Pijnenburg MW, Bakker EM, Lever S, Hop WC, De Jongste JC. High fractionalconcentration of nitric oxide in exhaled air despite steroid treatment inasthmatic children. Clinical & Experimental Allergy 2005;35(7):920–5.

[86] Dahgam S, Nyberg F, Modig L, Naluai AT, Olin AC. Single nucleotide polymor-phisms in the NOS2 and NOS3 genes are associated with exhaled nitric oxide. JMed Genet 2012;49:200–5.

[87] Storm van’s Gravesande K, Wechsler ME, Grasemann H, et al. Association of amissense mutation in the NOS3 gene with exhaled nitric oxide levels. Am JRespir Crit Care Med 2003;168:228–31.

[88] Ali M, Khoo SK, Turner S, Stick S, Le Souef P, Franklin P. NOS1 polymorphism isassociated with atopy but not exhaled nitric oxide levels in healthy children.Pediatr All Immunol 2003;14:261–5.

Paediatric Respiratory Reviews 16 (2015) 97–100

Mini-Sympoisum: Childhood asthma: The fuss and the future

Montelukast in paediatric asthma: where we are now and what stillneeds to be done?§

Andrew Bush *

Professor of Paediatrics and Head of Section (Paediatrics), Imperial College, Professor of Paediatric Respirology, National Heart and Lung Institute, Consultant

Paediatric Chest Physician, Royal Brompton Harefield NHS Foundation Trust

A R T I C L E I N F O

Keywords:

Asthma

Pre-school wheeze

Allergic rhinitis

Leukotriene

Inhaled corticosteroid

S U M M A R Y

Leukotriene receptor antagonists were introduced as an entirely new concept in asthma therapy, which

indeed they are. However, although an intellectually new concept, they have largely disappointed in

clinical practice. A small minority of school age asthmatics may respond better to these medications as

against inhaled corticosteroids as prophylactic therapy. In children not responding to low dose inhaled

corticosteroids, the best add-on therapy is salmeterol, but a small number respond better to

Montelukast. In pre-school wheeze, intermittent Montelukast may be an effective strategy in some

children who wheeze just with viral colds, but the clinical trial data are controversial. Pre-schoolers with

multiple trigger wheeze are probably best treated with inhaled corticosteroids. What is clear is that

clinically, a higher proportion of children are prescribed Montelukast than would be predicted from the

lterature to respond to the medication. No biomarker to predict response to Montelukast has reached

clinical practice, so N of 1 clinical trials should be performed. It is important not to leave children on

Montelukast if there is no convincing response to this treatment.

� 2014 Elsevier Ltd. All rights reserved.

EDUCATIONAL AIMS

The reader will come to appreciate that:

� The use of leukotriene receptor antagonists in asthma management is commonplace and likely excessive.� The clinical response to montelukast varies considerably and unpredictably in children, reinforcing the need for better biomarkers

in the management of asthma.� Those most likely to benefit from montelukast would appear to be younger, less atopic children with milder, in particular episodic

symptoms.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

INTRODUCTION

Leukotriene receptor antagonists (LTRAs) were (rightly) hailedas a whole new and novel class of asthma medications, distinctfrom the steroid based therapies and short and long acting b-2agonists. However there is all the difference in the world between

§ AB was supported by the NIHR Respiratory Disease Biomedical Research Unit at

the Royal Brompton and Harefield NHS Foundation Trust and Imperial College

London.* Department of Paediatric Respiratory Medicine, Royal Brompton Hospital,

Sydney Street, London SW3 6NP, UK. Tel.: +207-351-8232; fax: +207-351-8763.

E-mail address: a.bush@imperial.ac.uk.

http://dx.doi.org/10.1016/j.prrv.2014.10.007

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

an intellectually interesting concept and a therapeutic great leapforward. It is probably fair to say that the latter has not beendelivered. The aim of this manuscript is to review the currentpositioning of montelukast in the context of pre-school wheezeand childhood asthma; and, given the current uncertainties in thetreatment role of montelukast, to propose both mechanistic andclinical ways forward.

ROLE OF LTRAS IN SCHOOL AGE ASTHMA

A Cochrane review of every rigorously conducted, head to headcomparison between LTRAs and inhaled corticosteroids (ICS) camedown unequivocally for the superiority of ICS [1]. It is true that a

A. Bush / Paediatric Respiratory Reviews 16 (2015) 97–10098

so-called ‘real life study’ has reported equivalence [2]. However,such studies often lack rigour; how many children actually neededany treatment? By definition, two unnecessary treatments will beequally (un)efficacious; likewise, if patients are equally uncom-pliant with two different treatments, they are likely to beindistinguishable in terms of outcomes. Indeed, the notion thatadherence to a tablet is likely to be better than to an inhaler is, likeso many ideas about adherence, likely fallacious [3]. So forexample, one head to head comparison recruited 144 children aged6-17 years, of whom 17 dropped out, with mild-moderatepersistent asthma [4]. They used two different crossover regimes,of treatment with either fluticasone or montelukast, of eight weekseach. The end-point was an improvement of first second forcedexpired volume (FEV1) of more than 7.5%. Given the relatively poorutility of spirometry in school age asthma [5–7], this could havebeen criticised with the benefit of the retrospectoscope (as with somany studies with such a useful instrument!). However, theyreported similar findings with other endpoints, and also suggestedthat high exhaled nitric oxide (FeNO) might be a good predictor ofresponse to fluticasone [8]. Nonetheless, they reported that 55%did not respond to either medication, and 17% responded to both.Of those who differentially responded to a single medication, 5%responded to montelukast and 23% to fluticasone. Overall,24 children had a better FEV1 response with montelukast therapy,compared to 75 with fluticasone. The differential responses wererelated to biomarkers at baseline. A favourable response tofluticasone, as against non-responders, was associated with ahigh FeNO, higher blood eosinophil count, and elevated serum sIgEand eosinophilic cationic protein (ECP). A favourable fluticasoneresponse was also associated with a lower Methacholine PC20

(PC20meth) and worse spirometry. A favourable response tomontelukast, as against non-responders, was seen in youngerchildren, with a shorter disease duration; it is speculative thatthese might have been children with episodic viral wheeze (EVW).A differential better response to fluticasone as against montelukastwas associated with greater use of bronchodilators and a betterresponse, a greater FeNO and serum ECP, worse spirometry and alower PC20meth. Overall, fluticasone seemed to be the treatment ofchoice if spirometry was low and airway inflammation marked.The issue of biomarkers for differential response has been takenfurther in subsequent studies (below). What has not been tested inthis or any other study is whether this sort of differential responseis consistent within an individual over time.

The BADGER study addressed the important question of howbest to manage the asthmatic child who remains symptomaticdespite being prescribed (and inhaling!) 100 mcg fluticasone twicedaily [9]. The options tested in a triple crossover design wereadding either the long acting the b-2 agonist salmeterol, ormontelukast, or increasing the dose of ICS to 250 mcg bd. FeNO,PC20meth, beta-receptor polymorphisms and the asthma controltest (ACT) were used as prospective biomarkers to predictresponse. The salient features of the results were (a) that themost successful strategy was the addition of salmeterol; (b), anddisappointingly for the lovers of biomarkers, responders tosalmeterol were predicted by an ACT of >19/25; and (c) that formost children, the plateau of the dose response curve was at thesurprisingly low dose of 200 mcg/day. Although some childrenresponded to the addition of montelukast, it was clear that formany, this was not a successful strategy. Subsequent reports alsosuggested that salmeterol was the best add-on therapy in childrenwithout eczema. In children with eczema there were racialdifferences in optimal step-up therapy, although, as the authorsrightly conclude, the data are hypothesis-generating and need tobe replicated in another population [10]. They also showed thatimpulse oscillometry predicted a better FEV1 response tosalmeterol, but that there was a non-significant trend

(p = 0.053) to urinary LT-E4 levels predicting a better responseto montelukast. Again, these results have to be consideredhypothesis-generating and requiring confirmation [11].

In summary, montelukast in large groups of school age childrenis inferior to ICS as a first line preventer and inferior to LABA as add-on therapy. Clinical experience is that some individuals maybenefit, but we do not know how to select them prospectively.Clinical experience is also that many children are left onleukotriene receptor antagonists long term with no evidence ofbenefit, and no deterioration on stopping treatment.

ROLE OF LTRAS IN PRE-SCHOOL AGE WHEEZE

Numerous guidelines have stressed the phenotypic dissimi-larity between at least some pre-school wheezing syndromesand school age asthma [12,13]. Episodic (viral) wheeze (EVW)is defined as wheeze in association with a (usually) clinicallydiagnosed viral upper respiratory tract infection (URTI); itis not synonymous with any of the transient early wheezesyndromes.

Multi-trigger wheezers wheeze both with viral URTI and alsoother triggers between URTIs, such as excited behaviour andallergen exposure. It should be noted that it is not the same aspersistent wheeze. There is independent pathological support forthis classification; MTW but not EVW children have eosinophilicinflammation on airway biopsy [14] and MTW worse airflowobstruction and a higher FeNO [15]. It should be noted in passingthat these phenotypes may vary with time and the child should bere-evaluated regularly. In terms of treatment of pre-schoolwheeze, we have no disease modifying therapies; at least threegood studies [16–18] have shown that early institution of ICStreatment even in those with a higher risk of developing asthma(positive modified asthma predictive index [19]) does not reducethe risk of the subsequent development of asthma, so treatmentshould be based on symptoms.

Three studies have addressed the question as to whetherintermittent montelukast is effective in treating pre-schoolwheeze [20–22]. The PREEMPT study [20] compared intermittentmontelukast with placebo (>100 children and >300 exacerba-tions/group). The benefit was in the youngest children, and therewas around a one third reduction in the time the child wasremoved from a childcare facility and the time the carer was offwork. A North American study [21] in 238 pre-school childrencompared intermittent nebulised budesonide (the only aerosolisedsteroid permitted by the FDA), intermittent montelukast andplacebo, given at the time of a viral-induced exacerbation. Therewere minor and equivalent benefits for montelukast andbudesonide over placebo, but the end-points were rather softand the results not dramatic. The largest study of all [22] recruitedintermittently wheezing children age 6/12 to 5 years; 589 weretreated with daily Montelukast, 591 with intermittent Montelu-kast, and 591 with placebo. The primary end-point was episodesculminating in an asthma attack. There were a mean of4 exacerbations/child, and therefore more than 2000 exacerba-tions/group. There was no improvement in the primary end-points,but statistically significant numerical improvements in some 2ry

end-points; however this must be considered a negative study. TheALOX study [23] is the subject of another manuscript in this mini-symposium, but does not alter my conclusions as to the role ofmontelukast in pre-school wheeze.

So where if anywhere should we position montelukast in thetreatment of pre-school wheeze? Clearly it will not work for allchildren, but clinical experience is that there are a sub-group forwhom this treatment is dramatically effective; and it should benoted that there are few dramatic therapeutic successes in thiscontext. I suggest that, whereas ICS are first line preventive

A. Bush / Paediatric Respiratory Reviews 16 (2015) 97–100 99

treatment for MTW, especially if the child is atopic [24],montelukast can be positioned as first line ‘preventive treatment’in EVW in pre-school children, provided that preventive treatmentis justified by the symptoms. I would trial montelukast intermit-tently with viral colds as first line therapy. If that does not work, Iwould next go to continuous therapy, especially in children whohave a pattern of very rapid deterioration. I always use a three stepprotocol to ensure that any apparent improvement is not due to thenatural history of the disease (Text Box). Although in the main,montelukast is a very safe medication, it is essential to warnparents about potential behavioural side-effects. These are rare butalarming, but fortunately recover if treatment is stopped.

HOW DO WE MOVE FORWARD?

It is clear that Montelukast has at best a role in a minority ofasthmatics at any age. This was confirmed in an adult study of reallysevere asthma [25]; 100 adult asthmatics were recruited. They wereprescribed ICS 250-4000mcg (Fluticasone equivalent) and the datawere analysable in 91. The protocol was 2 weeks Montelukast/Placebo, followed by 2 weeks wash out, and then 2 weeks Placebo/Montelukast in random order. There was an independent statisticalanalysis. There was absolutely no discernable benefit. Nonetheless,every series reporting really severe asthma includes numerouspatients who continue to take Montelukast. So for example, in theSARP study, a minimum of 67% of their cluster 5 (fixed airflowobstruction) and 54% of cluster 3 (obese, very late onset) were usingMontelukast [26]! In two papers from our group [27,28] 50/102(49%) and 31/53 (58%) were prescribed Montelukast. It isinconceivable from the literature that all these patients are gettingbenefit. Anecdotally, many patients are able to stop Montelukastwithout any deterioration. More likely, equivocal transient benefitwas noted and the physician did not have the courage to discontinuetherapy in a symptomatic patient. It would seem to be good practiceto use the six week, three stage protocol above in clinical practice;unfortunately, the lack of availability of placebo precludes this.Perhaps legislators should consider mandating the availability ofplacebos and this sort of trial for some medications in the future.

If the ideal of a placebo controlled, therapeutic trial is notattainable, what about biomarkers to predict response? Despite anumber of attempts in big groups, no biomarker has met the twincriteria of being validated in a second cohort and reaching routineclinical practice. A re-analysis of the PACT trial reported on191 children, median age 9.5 years, who were prescribed eitherfluticasone 100 mcg bd or Montelukast 5 mg od [29]. Potentialpredictors which were studied were age at onset of symptoms anddiagnosis of asthma; presence of atopy; IgE levels; bronchialhyper-responsiveness; blood eosinophil count and ECP; andurinary LTE4. The authors concluded that the best predictors of

Text Box. Three step protocol for prophylactic treatment of

pre-school wheeze

� Step 1: commence montelukast 4-5 mg (depending on age)

once daily

� Step 2: Discontinue after 6-8 weeks

- If still symptomatic, further treatment futile

- If symptom free, could be either therapeutic success, or

natural history of the condition

� Step 3: Only re-start montelukast if symptoms recur after

step 2

ICS response ‘might’ be a history of parental asthma, an increasedFeNO, a low PC20 and previous ICS use. This was taken further in astudy of 318 asthmatic children enrolled in CLIC & PACT [30], bothas a combined cohort and separate analyses of the two studies. Therelationship between urinary LTE4:FeNO ratio and spirometry andasthma control days was determined; an elevated LTE4:FeNO ratio(analysing both cohorts together) associated with better Mon-telukast response for FEV1 and asthma control days. The authorsfound that those with a ratio >75th percentile were more likely tobe younger, female, less atopic, and have a higher highermethacholine PC20. However, this has yet to be validated inanother large group, which needs to be done before it can berecommended to be utilised in clinical practice.

The suggestion has been made that children with asthma andallergic rhinitis as a co-morbidity may respond better tomontelukast than children with asthma alone. It is clear thatallergic rhinitis should be treated on its own merits, since it causesconsiderable morbidity [31]. Furthermore, in terms of upperairway disease, Montelukast may have a role in treating mild sleepdisordered breathing and obstructive sleep apnoea [32]. However,there is no evidence that allergic rhinitis as a co-morbidity predictsa better response to Montelukast.

Adult smokers who are asthmatics are a group traditionallyignored in randomised controlled trials, for fear of confusion withCOPD. There are many studies suggesting that active smoking maycause steroid resistance [33–35]. So could asthmatics who smokebe an attractive group to treat with Montelukast? 44 non-smokersand 39 light smokers age 20-50, with known asthma wererandomly assigned to beclomethasone 200 mcg bd or Montelukast10 mg od for 8 weeks. The groups were well matched at baselinefor lung function, airway inflammation measured using inducedsputum, and BHR [36]. The primary endpoint was change in pre-bronchodilator FEV1, secondary outcomes were morning andevening peak expiratory flow rate, PC20 methacholine, symptoms,Quality of Life, and sputum cytology, ECP and tryptase. In terms ofthe primary end-point, there was no improvement in FEV1 in eithergroup with Montelukast, nor in the smoking asthmatics usingbeclomethasone; the non-smoking asthmatics, as expected,improved their FEV1. The results of the secondary end-pointswere challenging. Despite the change in FEV1, morning peak flowonly improved in the asthmatic smokers who were treated withmontelukast. There were no significant chanes in either group witheither treatment for sputum eosinophils or PC20 methacholine.This study suggests that Montelukast may be of benefit inasthmatics who smoke; however, this needs confirmation.

Are children who are exposed passively to tobacco smoke, an alltoo common scenario [37], also steroid resistant? The histonedeacetylases (HDAC) acts on nuclear chromatin to control geneexpression. HDAC2 is a pre-requisite for corticosteroids to switch offactivated inflammatory genes. We studied 19 school age childrenundergoing fibreoptic bronchoscopy for severe, therapy resistantasthma [38], of whom nine were exposed to environmental tobaccosmoke. HDAC2 protein and HDAC2 activity were reduced, whereasHDAC1 and total HDAC activity was unchanged. One couldspeculate that passive smoke exposure may be a determinant ofresponse to Montelukast in children. However, this needs to betested prospectively; and the best treatment of passive smokingrelated asthma is not Montelukast, but smoking cessation.

In summary, we do not have biomarkers to predict the responseto Montelukast. The suggestion is that the responders to ICS maybe more atopic, and have more eosinophilic inflammation thannon-responders. Whether this corresponds to the so-called TH2-hi[39] group in adults remains to be seen. Although this findingraises interesting possibilities for research, currently inflammationand airway reactivity to predict response to therapy in theindividual child is not clinically useful.

A. Bush / Paediatric Respiratory Reviews 16 (2015) 97–100100

SUMMARY AND CONCLUSIONS

Leukotriene receptor antagonists are an exciting novel class ofagents from an academic perspective, but at best it is clear they arebeneficial for a small minority of pre-school wheezers and schoolage asthmatics. It is also clear they are prescribed to far morepeople than actually benefit. We need robust and clinicallyapplicable biomarkers of response to Montelukast; one approachwould be to select patients who are dramatic responders toleukotriene receptor antagonists and a group of total non-responders, and compare their genetics in minute detail, includingSNPs and gene transcription protocols. Pending such biomarkers,we are compelled to do N of 1 clinical trials, which would be greatlyfacilitated by having placebos, so that neither paediatrician norfamily would know whether active or placebo was beingadministered. Above all, it is essential critically to assess responseto these agents, and prevent children being left on leukotrienereceptor antagonists by default.

FUTURE RESEARCH DIRECTIONS

� Better genetic characterisation of children who respond well tomontelukast in comparison to those children who don’t.� The development and validation of accurate and reliable

biomarkers of asthma control in children.

References

[1] Chauhan BF, Ducharme FM. Anti-leukotriene agents compared to inhaledcorticosteroids in the management of recurrent and/or chronic asthma inadults and children. Cochrane Database Syst Rev 2012 May 16;5. CD002314.

[2] Price D, Musgrave SD, Shepstone L, Hillyer EV, Sims EJ, Gilbert RF, Juniper EF,Ayres JG, Kemp L, Blyth A, Wilson EC, Wolfe S, Freeman D, Mugford HM,Murdoch J, Harvey I. Leukotriene antagonists as first-line or add-on asthma-controller therapy. N Engl J Med 2011;364:1695–707.

[3] Carter ER, Ananthakrishnan M. Adherence to montelukast versus inhaledcorticosteroids in children with asthma. Pediatr Pulmonol 2003;36:301–4.

[4] Szefler SJ, Phillips BR, Martinez FD, et al. Characterization of within-subjectresponses to fluticasone and montelukast in childhood asthma. J Allergy ClinImmunol 2005;115:233–42.

[5] Bacharier LB, Strunk RC, Mauger D, et al. Classifying asthma severity inchildren: mismatch between symptoms, medication use, and lung function.Am J Respir Crit Care Med 2004;170:426–32.

[6] Expert Panel Report 3 (EPR-3): guidelines for the diagnosis and management ofasthma-summary report 2007. J Allergy Clin Immunol 2007; 120 (Suppl): S94-138.

[7] van Dalen C, Harding E, Parkin J, et al. Suitability of forced expiratory volume in1 second/forced vital capacity vs percentage of predicted forced expiratoryvolume in 1 second for the classification of asthma severity in adolescents.Arch Pediatr Adolesc Med 2008;162:1169–74.

[8] Zeiger RS, Szefler SJ, Phillips BR, et al. Response profiles to fluticasone andmontelukast in mild-to-moderate persistanet childhood asthma. J Allergy ClinImmunol 2006;117:45–52.

[9] Lemanske Jr RF, Mauger DT, Sorkness CA, et al. Childhood Asthma Research andEducation (CARE) Network of the National Heart, Lung, and Blood Institute.Step-up therapy for children with uncontrolled asthma receiving inhaledcorticosteroids. N Engl J Med 2010;362:975–85.

[10] Malka J, Mauger DT, Covar R, et al. Eczema and race as combined determinantsfor differential response to step-up asthma therapy. J Allergy Clin Immunol2014; epub.

[11] Rabinovitch N, Mauger DT, Reisdorph N, et al. Predictors of asthma control andlung function responsiveness to step 3 therapy in children with uncontrolledasthma. J Allergy Clin Immunol 2014;133:1450–6.

[12] Brand PL, Baraldi E, Bisgaard H, Boner AL, Castro-Rodriguez JA, Custovic A, deBlic J, de Jongste JC, Eber E, Everard ML, Frey U, Gappa M, Garcia-Marcos L, GriggJ, Lenney W, Le Souef P, McKenzie S, Merkus PJ, Midulla F, Paton JY, Piacentini G,Pohunek P, Rossi GA, Seddon P, Silverman M, Sly PD, Stick S, Valiulis A, vanAalderen WM, Wildhaber JH, Wennergren G, Wilson N, Zivkovic Z, Bush A.Definition, assessment and treatment of wheezing disorders in preschoolchildren: an evidence-based approach. Eur Respir J 2008;32:1096–110.

[13] Brand PL, Caudri D, Eber E, Gaillard EA, Garcia-Marcos L, Hedlin G, Henderson J,Kuehni CE, Merkus PF, Pedersen S, Valiulis A, Wennergren G, Bush A. Classifi-cation and pharmacological treatment of preschool wheezing: changes since2008. Eur Respir J 2014;43:1172–7.

[14] Sonnappa S, Bastardo CM, Wade A, Saglani S, McKenzie SA, Bush A, Aurora P.Symptom-pattern and pulmonary function in preschool wheezers. J AllergyClin Immunol 2010;126:519–26.

[15] Sonnappa S, Bastardo CM, Saglani S, Bush A, Aurora P. Relationship betweenpast airway pathology and current lung function in preschool wheezers. EurRespir J 2011;38:1431–6.

[16] Bisgaard H, Hermansen MN, Loland L, et al. Intermittent inhaled corticoste-roids in infants with episodic wheezing. N Engl J Med 2006;354:1998–2005.

[17] Guilbert TW, Morgan WJ, Zeiger RS, et al. Long-term inhaled corticosteroids inpreschool children at high risk for asthma. N Engl J Med 2006;354:1985–97.

[18] Murray CS, Woodcock A, Langley SJ, et al. Secondary prevention of asthma bythe use of inhaled fluticasone dipropionate in wheezy Infants (IWWIN):double-blind, randomised controlled study. Lancet 2006;368:754–62.

[19] Guilbert TW, Morgan WJ, Zeiger RS, Bacharier LB, Boehmer SJ, Krawiec M,Larsen G, Lemanske RF, Liu A, Mauger DT, Sorkness C, Szefler SJ, Strunk RC,Taussig LM, Martinez FD. Atopic characteristics of children with recurrentwheezing at high risk for the development of childhood asthma. J Allergy ClinImmunol 2004;114:1282–7.

[20] Robertson CF, Price D, Henry R, et al. Short-course montelukast for intermit-tent asthma in children: a randomized controlled trial. Am J Respir Crit CareMed 2007;175:323–9.

[21] Bacharier LB, Phillips BR, Zeiger RS, et al., CARE Network. Episodic use of aninhaled corticosteroid or leukotriene receptor antagonist in preschool childrenwith moderate-to-severe intermittent wheezing. J Allergy Clin Immunol2008;122:1127–35.

[22] Valovirta E, Boza ML, Robertson CF, et al. Intermittent or daily montelukastversus placebo for episodic asthma in children. Ann Allergy Asthma Immunol2011;106:518–26.

[23] Nwokoro C, Pandya H, Turner S, Eldridge S, Griffiths CJ, Vulliamy T, Price D,Sanak M, Holloway JW, Brugha R, Koh L, Dickson I, Rutterford C, Grigg J.Intermittent montelukast in children aged 10 months to 5 years with wheeze(WAIT trial): a multicentre, randomised, placebo-controlled trial. Lancet RespirMed 2014;2:796–803.

[24] Castro-Rodriguez JA, Rodrigo GJ. Efficacy of inhaled corticosteroids in infantsand preschoolers with recurrent wheezing and asthma: a systematic reviewwith meta-analysis. Pediatrics 2009;123:e519–25.

[25] Robinson DS, Campbell D, Barnes PJ. Addition of leukotriene antagonists totherapy in chronic persistent asthma: a randomised double-blind placebo-controlled trial. Lancet 2001;357:2007–11.

[26] Moore WC, Fitzpatrick AM, Li X, Hastie AT, Li H, Meyers DA, Bleecker ER.Clinical heterogeneity in the severe asthma research program. Annal AmThorac Soc 2013;10Suppl:S118–22.

[27] Bossley CJ, Saglani S, Kavanagh C, Payne DN, Wilson N, Tsartsali L, Rosenthal M,Balfour-Lynn IM, Nicholson AG, Bush A. Corticosteroid responsiveness andclinical characteristics in childhood difficult asthma. Eur Respir J 2009;34:1052–9.

[28] Bossley C, Fleming L, Gupta A, Regamey N, Frith J, Oates T, Tsartali L, Lloyd C,Bush A, Saglani S. Pediatric severe asthma is characterized by eosinophiliaand remodeling without TH2 cytokines. J Allergy Clin Immunol 2012;129:974–82.

[29] Knuffman JE, Sorkness CA, Lemanske Jr RF, et al. Childhood Asthma Researchand Education Network of the National Heart, Lung, and Blood Institute.Phenotypic predictors of long-term response to inhaled corticosteroid andleukotriene modifier therapies in pediatric asthma. J Allergy Clin Immunol2009;123:411–6.

[30] Rabinovitch N, Graber NJ, Chinchilli VM, et al. Childhood Asthma Research andEducation Network of the National Heart, Lung, and Blood Institute. Urinaryleukotriene E4/exhaled nitric oxide ratio and montelukast response in child-hood asthma. J Allergy Clin Immunol 2010;126:545–51.

[31] Sheikh A, Khan-Wasti S, Price D, Smeeth L, Fletcher M, Walker S. Standardizedtraining for healthcare professionals and its impact on patients with perennialrhinitis: a multi-centre randomized controlled trial. Clin Exp Allergy2007;37:90–9.

[32] Goldbart AD, Greenberg-Dotan S, Tal A. Montelukast for children with ob-structive sleep apnea: a double-blind, placebo-controlled study. Pediatrics2012;130:e575–80.

[33] Chalmers G, Macleod K, Little S, Thomson L, McSharry C, Thomson N. Influenceof cigarette smoking on inhaled corticosteroid treatment in mild asthma.Thorax 2002;57:226–30.

[34] Tomlinson J, McMahon A, Chaudhuri R, Thompson J, Wood S, Thomson N.Efficacy of low and high dose inhaled corticosteroid in smokers versus non-smokers with mild asthma. Thorax 2005;60:282–7.

[35] Mark Spears, Charles McSharry, Rekha Chaudhuri, Christopher J. Weir, Carl deWet, Neil C. Thomson Smoking in Asthma Is Associated with Elevated Levels ofCorticosteroid Resistant Sputum Cytokines—An Exploratory Study. PLoS One.2013; 8: e71460.

[36] Lazarus SC, Chinchilli VM, Rollings NJ etal, National Heart Lung, Blood Insti-tute’s Asthma Clinical Research Network. Smoking affects response to inhaledcorticosteroids or leukotriene receptor antagonists in asthma. Am J Respir CritCare Med 2007;175:783–90.

[37] Bracken MB, Fleming L, Hall P, Van Stiphout N, Bossley CJ, Biggart E, WilsonNM, Bush A. The importance of nurse led home visits in the assessment ofchildren with problematic asthma. Arch Dis Child 2009;94:780–4.

[38] Kobayashi Y, Bossley C, Gupta A, Akashi K, Tsartsali L, Mercado N, Barnes PJ,Bush A, Ito K. Passive smoking impairs histone deacetylase-2 in children withsevere asthma. Chest 2014;145:305–12.

[39] Woodruff PG, Modrek B, Choy DF, Jia G, Abbas AR, Ellwanger A, Koth LL, ArronJR, Fahy JV. T-helper type 2-driven inflammation defines major subphenotypesof asthma. Am J Respir Crit Care Med 2009;180:388–95.

Paediatric Respiratory Reviews 16 (2015) 101–107

Mini-Sympoisum: Childhood asthma: The fuss and the future

Personalized medicine in children with asthma

Marielle W. Pijnenburg 1,*, Stanley Szefler 2

1 Department of Paediatrics/ Paediatric Respiratory Medicine, Erasmus Medical Centre – Sophia Children’s Hospital, Rotterdam, The Netherlands2 The Breathing Institute / Pulmonary Medicine, Department of Pediatrics, Children’s Hospital Colorado; University of Colorado Denver School of Medicine,

Aurora (CO), USA

EDUCATIONAL AIMS THE READER WILL COME TO APPRECIATE THAT:

� asthma is a heterogeneous disease with a huge variability in asthma phenotypes, in genetic background of patients, age, severity ofasthma, risk factors and co-morbidities which may warrant different and more personalized treatment and monitoring approaches� several patient characteristics, lung function parameters and biomarkers have been shown to predict treatment response to

inhaled corticosteroids (ICS), montelukast or long-acting beta-agonists or to predict successful reduction of ICS� the number of genes identified for the various asthma drug response phenotypes is small and limits the use of pharmacogenetics in

asthma treatment to date� e-health may allow for personalized treatment and monitoring; studies on the most feasible interventions in individual patients

are needed

A R T I C L E I N F O

Keywords:

Asthma

Children

Personalized medicine

Pharmacogenetics

E-health

Biomarkers

S U M M A R Y

Personalized medicine for children with asthma aims to provide a tailored management of asthma, which

leads to faster and better asthma control, has less adverse events and may be cost saving.

Several patient characteristics, lung function parameters and biomarkers have been shown useful in

predicting treatment response or predicting successful reduction of asthma medication.

As treatment response to the main asthma therapies is partly genetically determined, pharmaco-

genetics may open the way for personalized medicine in children with asthma. However, the number of

genes identified for the various asthma drug response phenotypes remains small and randomized

controlled trials are lacking.

Biomarkers in exhaled breath or breath condensate remain promising but did not find their way from

bench to bedside yet, except for the fraction of exhaled nitric oxide.

E-health will most likely find its way to clinical practice and most interventions are at least non-inferior

to usual care. More studies are needed on which interventions will benefit most individual children.

� 2014 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

INTRODUCTION

Asthma is the leading cause of chronic disease in children inthe western world and affects approximately 1 out of 10 children[1]. Although effective medications such as the inhaled corticoste-roids (ICS) and updated guidelines on asthma in children areavailable, a substantial proportion of children (40-70%) has only

* Corresponding author. Department of Paediatrics/ Paediatric Respiratory

Medicine, Erasmus Medical Centre – Sophia Children’s Hospital, Wijtemaweg 12,

3015 CN Rotterdam, The Netherlands. Tel.: +31 10 7036263; fax: +31 10 7036811.

E-mail addresses: m.pijnenburg@erasmusmc.nl (M.W. Pijnenburg),

Stanley.Szefler@childrenscolorado.org (S. Szefler).

http://dx.doi.org/10.1016/j.prrv.2014.10.003

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

partly or even poorly controlled disease [2–7]. Guidelines suggest aone size fits all approach with a step up and step down schemebased on asthma control for all patients [3,4]. However, asthma is aheterogeneous disease with a huge variability in asthma pheno-types, in genetic background of patients, age, severity of asthma,risk factors and co-morbidities that may warrant different andmore personalized treatment and monitoring approaches [8]. Forexample, while some children with mild to moderate asthma willbenefit from treatment with ICS, others may even deteriorate onICS and montelukast might be the preferred option. Also, in step3 asthma treatment it would be very helpful to have more insighton which children will benefit most from the different treatmentoptions in order to manage them with the best possible results and

Table 1Step 2 treatment: ICS vs montelukast

Favors ICS response Favors LTRA response

FEV1 as outcome:

- Higher SABA use

- higher bronchodilator response

- higher FeNO

- higher ECP levels

- lower methacholine PC(20)

- lower pulmonary function values

Asthma control days as outcome:

- higher FeNO

Asthma control and lung function as outcome:

- parental history of asthma

- higher FeNO

- low PC20 values

- history of ICS use

FEV1 as outcome:

- Younger age

- Shorter duration of asthma

Asthma control and lung

function as outcome:

- urinary leukotriene

E4/ FeNO ratio

Table 2Predictive factors for step 3 treatment (doubling ICS, adding LABA, adding LTRA)

Predictor Effect

- baseline FEV1

- bronchodilator response

- baseline FeNO

- (C-) ACT > 19

- methacholine PC20 values

- genotype at position 16 of the b2-adrenergic receptor

- no history of eczema

- higher IOS reactance area

- race

Not predictive

Not predictive

Not predictive

Favors LABA

Not predictive

Not predictive

Favors LABA

Favors LABA

Not predictive*

* In children with eczema, black children responded better to LABA step-up,

white Hispanics to LTRA step-up.

M.W. Pijnenburg, S. Szefler / Paediatric Respiratory Reviews 16 (2015) 101–107102

the least risk of adverse effects while being cost effective. Althoughin most children symptom based-management may be sufficient,others will need more intense monitoring programs to achievebetter outcomes [9]. Additionally, patient preferences and theirindividual goals of treatment may play an important role inchoosing the most appropriate treatment [10–14].

Therefore the aim of personalized medicine for children withasthma is to provide a tailored treatment and monitoring strategy,which is more safe, leads to faster asthma control, has less adverseevents and may be cost-saving. This review aims to summarize thecurrent state and future perspectives on personalized medicine inchildren with asthma.

PERSONALIZED VERSUS STRATIFIED TREATMENT

Personalized medicine is the customization of health caretailored to the individual; it uses new technology or discovery toenable a level of personalization not previously feasible orpractical. For example, in cancer therapy, genetic determinantsof the tumor determine which chemotherapeutic drugs or whichadjuvant therapy should be used for the best survival with the leastrisk of side effects. In other words personalized medicine aims toidentify biomarkers or genetic features of patients most likely torespond to a medication. On the other hand, these biologic markerscould be associated with little likelihood of response or even moreconcerning a likelihood of an adverse effect.

Personalized medicine should be balanced against stratifiedmedicine, which classifies individuals into subpopulations thatdiffer in their susceptibility to a particular disease or their responseto a particular treatment. For example, in comparison to children ofnormal weight, obese children are less likely to respond to ICS iflung function and exacerbations are the outcomes [15]. In asthmatreatment frequently stratified medicine is used, althoughpharmacogenetics makes personalized medicine within our reach.

Stratified medicine may be particularly useful in predictingtreatment response.

PREDICTING TREATMENT RESPONSE

Inhaled corticosteroids versus montelukast

Several studies addressed the question whether response to treat-ment may be predicted by clinical markers, lung function, biomarkersand/or genetic polymorphisms. Anticipating treatment responsemay improve symptoms more quickly and prevent side effects oftreatment in susceptible individuals. However, one should be awarethat phenotypic or genetic predictors of long-term treatmentresponse depend on the definition of outcome, like Forced ExpiratoryVolume in 1 second (FEV1) or asthma control days [16,17].

In children with persistent asthma who require step 2 asthmatreatment all guidelines prefer ICS over leukotriene receptorantagonists (LTRA), however, some children might benefit morefrom a LTRA [18]. In a cross over trial in 126 children with mild-moderate asthma, where response to ICS or LTRA was defined as animprovement of at least 7.5% predicted in FEV1, most children had adifferential response [19]. Children were more likely to respondbetter to fluticasone than to montelukast if they used morebronchodilators, had higher bronchodilator response, higher frac-tional exhaled nitric oxide (FeNO) levels, higher eosinophil cationicprotein (ECP) levels or lower methacholine PC20 and pulmonaryfunction values. Younger age and shorter duration of asthmawere associated with a more favorable outcome on montelukast[19]. With asthma control days as a primary outcome, FeNO wasboth a predictor and a response indicator in discriminating thedifference in treatment response between fluticasone and mon-telukast [20]. In a second study, parental history of asthma, elevated

FeNO, low PC20 values, or a history of ICS use predicted better long-term clinical and lung function outcomes with ICS compared to LTRA[21]. In contrast, the ratio between urinary leukotriene E4 to FeNOpredicted a better response in FEV1 and asthma control days ofmontelukast over fluticasone in children with mild-moderateasthma [22]. (Table 1)

As needed versus daily ICS

Four studies assessed whether daily ICS is superior to as neededICS use in children with mild asthma, however none of themstudied predictors of successful response to as needed treatment ascompared to daily ICS [23–26].

Step 3 treatment

In children who are not well controlled on low dose ICS, step upin treatment may be required and 3 potential options are available:doubling the dose of ICS, adding a long acting beta-2 agonist (LABA)or adding a LTRA. If we were able to pick out the best option foreach individual child, over – and under-treatment might beavoided and asthma control established quicker. In a crossoverstudy in 182 children with uncontrolled asthma, the 3 treatmentoptions were compared and the vast majority of patients showed adifferential response to the treatment options [27]. Neithermethacholine PC20 values, or FeNO, or genotype at position16 of the b2-adrenergic receptor, nor FEV1 or bronchodilatorresponse (post-hoc analysis) predicted to which treatment optionpatients would respond best. (Table 2) However, higher baselinescores on the Asthma Control Test (ACT) or childhood AsthmaControl Test (C-ACT) predicted a greater probability that the bestresponse would be to LABA step-up [27]. In a post hoc multivariateanalysis higher impulse oscillometry reactance area predictedhigher FEV1 response to LABA add on compared to ICS step up

M.W. Pijnenburg, S. Szefler / Paediatric Respiratory Reviews 16 (2015) 101–107 103

[28]. Also, children without a history of eczema were more likely torespond better to LABA step up and race appeared to differentiateresponders to ICS from responders to LTRA [29]. However, thecomplex interaction between genetic variation, race and responseto treatment needs more research.

In a parallel-group double-blind randomized controlled trialcomparing LABA add on versus doubling the dose of ICS in158 children with uncontrolled asthma, FEV1 and FeNO were notable to predict response to 1 of the 2 treatments [30].

Cluster analysis may give more insight into which groups ofpatients may benefit more from one treatment over the other [31].

Stepping down treatment

Except for factors predicting which step up in treatment is likelyto be most successful, predictors of successful stepping downtreatment may be equally important. Three studies addressed thequestion regarding safe reduction of ICS and its associationbetween biomarkers or lung function. Increased FeNO andincreased percentage of eosinophils in induced sputum, but notbronchial responsiveness, predicted a failure to tolerate reductionof ICS in 40 asthmatic children [32,33]. Similarly, FeNO measured2-4 weeks after withdrawal of ICS in asymptomatic childrenpredicted relapse of asthma [34]. In children in whom asthmamedications were stepped down, either according to guidelines oron the initiative of patient and/or caregivers, age, sex, tobaccosmoke exposure and FEV1 did not predict successful steppingdown of asthma medications [35]. However, stepping down in fallhad a greater failure rate compared to stepping down in otherseasons; being guideline eligible for step down predictedsuccessful step down in univariate analyses.

New predictors are needed to assess which children may besafely stepped down in asthma medications.

Biologicals

In children with uncontrolled persistent severe asthma biologi-cals like omalizumab (anti-IgE-antibodies), lebrikizumab (anti-Interleukine-13) and mepolizumab (anti-Interleukine-5) may be(future) options. These new drugs carry a high burden to the patientin regards to convenience of administration and are expensive.Therefore patient preference may be even more important forselecting these therapies. Children with high FeNO levels, bloodeosinophils, and high body mass index were more likely to respondto omalizumab and the effect of lebrikizumab on lung functionwas much higher in asthmatic adults with high periostin levelscompared with those with low periostin levels [36,37].

PHARMACOGENETICS

Although current asthma treatment is effective in mostpatients, response to treatment is heterogeneous. For the threemain asthma therapies (ICS, beta-2 agonists and LTRA) responseis partly genetically determined, although the number of genesidentified for the various asthma drug response phenotypesremains small. Pharmacogenetics is the study of the genetic basisfor interindividual differences in the reponse to drugs with the goalto identify patients with good response and reduce side effects andcost of medications.

The Arg16Gly single nucleotide polymorphism has been linkedto the response to treatment with short-acting beta-agonists(SABA) and LABA and is the most extensively studied polymor-phism in children with asthma. The effect of ADRB2 genotypeon treatment response to beta-agonists seems more evidentfor children compared to adults with asthma [38–40] and theArg16Arg polymorphism has been associated with an impaired

therapeutic effect of SABA and LABA [38–42], although data areconflicting [43–46].

In asthmatic children with the homozygous Arg16 genotypeadd on therapy with montelukast showed less school absences,less symptoms and better quality of life than add-on with LABA[38]. However, pharmacogenetic studies have only recently beenundertaken in children with asthma and randomized controlledtrials are needed to study if genotyping of the ADRB2 polymor-phisms will be of any clinical benefit.

Two other genes, ARG1 and GSNOR, have been associated withbeta-2 agonist response [47,48]. Genotype variation in GSNORwas associated with a decreased response to albuterol in AfricanAmerican children with a severe asthma attack, however replica-tion in other cohorts is needed. The ARG1 gene encodes arginaseand has been linked to bronchodilator response [48].

ICS are the cornerstone of asthma treatment and most childrenwith mild to moderate asthma may be well controlled on low doseICS [2–5]. However, some children seem to be steroid-resistantwhilst others show adverse effects, such as decreased growthvelocity and decreased bone density. Several genes have beenidentified as associated with ICS response on lung function, airwayhyperresponsiveness or exacerbations. A polymorphism of thecorticotrophin-releasing hormone receptor 1 (CRHR1) has beenrelated to improved lung function with ICS treatment [49]. Simi-larly, polymorphisms in the TBX21 gene were associated withimprovement in airway hyperresponsiveness [50]. Other genesinvolved in ICS response are GLCC11 gene (rs37973 associatedwith change in lung function) [51], FCER2 (T2206C, associatedwith exacerbations) [52,53], STIP 1 gene (rs4980524, rs6591838associated with change in lung function) and HDAC1 and HDAC2(rs17411981, associated with change in lung function). [54]Recently, a novel SNP, rs10044254, was found to be associatedwith improved symptomatic response to ICS in 2 independentpediatric cohorts, but not in an adult cohort [55].

The LTRAs are mostly used as add on treatment in patients withmore severe symptoms [3,4]. Leukotrienes are derived fromarachidonic acid, a reaction that is catalyzed by 5-lipoxygenase.Variants in ALOX5 (the gene encoding for 5-lipoxygenase), LTC4S(encoding for leukotriene C4 synthase), LTA4H (LT A4 hydrolase)and MRP1 (multidrug resistance protein 1) were associated withchanges in FEV1 and/or peak expiratory flow rates and exacerba-tion rates [56–60].

Variants in the SLCO2B1 gene, which is involved in metabolismof drugs in the liver, affect montelukast plasma levels [61].

Randomized controlled trials are needed to determine if asthmaoutcomes in individual children can be improved by usingpharmacogenetic predictors of treatment response to one of theasthma drugs.

BIOMARKERS

Biomarkers may play a role in personalized medicine byimproving diagnosis, monitoring of asthma, predicting treatmentresponse and to ‘phenotype’ patients into clusters with similarcharacteristics.

Fraction of exhaled nitric oxide

The most extensively studied biomarker is the fraction ofexhaled nitric oxide (FeNO), which can be easily and reproduciblymeasured in children from the age of 4 [62]. FeNO has beensuggested as a marker of eosinophilic airways inflammation, iselevated in steroid-naıve asthmatics and decreases after treatmentwith ICS [63–66]. As discussed previously, patients with elevatedFeNO levels are more likely to respond to ICS as compared tomontelukast and FeNO may predict successful reduction of ICS

M.W. Pijnenburg, S. Szefler / Paediatric Respiratory Reviews 16 (2015) 101–107104

[19–21,32–34]. The role of FeNO in monitoring children withasthma has been less clear and based on current evidence FeNO isnot useful for routine monitoring of children with asthma[67]. However, there might be a place for FeNO monitoring ICSresponse and adherence in subgroups of children with difficult oruncontrolled asthma, obese children or children with unexplainedasthmatic symptoms [68,69]. Also, the use of personalized FeNOcut-offs for guiding asthma treatment has not been established.

Sputum eosinophils

Analysis of induced sputum allows for assessing inflammatorycells such as eosinophils and neutrophils as well as inflammatorymediators. Sputum eosinophils may predict failed reduction of ICSin stable patients [32]. In contrast to adults, in children with severeasthma, tailored intervention based on sputum eosinophils did notreduce asthma exacerbations or improve asthma control whencompared to usual care [70]. Similarly, sputum eosinophils did notpredict response to systemic steroids in children with severeasthma [71].

Exhaled breath/ exhaled breath condensate

The analysis of volatile and non-volatile substances in exhaledbreath (EB) and exhaled breath condensate (EBC) for diagnosis andmanagement of respiratory diseases has raised great interest as anon-invasive method to assess airway inflammation and definedisease phenotypes. EB contains numerous volatile organic com-pounds (VOCs), which can discriminate asthmatic from healthychildren, and atopics from non-atopics [72–75]. A combination of6 VOCs was able to predict exacerbations in asthmatic children[76]. Although the results of EB analysis seem promising, eachindividual study demonstrates a different superior set of VOCs, andthere clearly is a need for external validation. The interpretation ofresults is seriously hampered by the lack of standardization. Also,further research is needed on the effects of disease severity and ICSon VOCs in children with asthma. Hence, the role of VOCs in EB inasthma monitoring and personalized medicine is still unclear.

Recently, analysis of exhaled breath condensate (EBC) inpaediatric asthma has been summarized by Thomas et al. [77]Most studies have focused on pH, hydrogen peroxide (H2O2) and8-isoprostane. Leukotrienes, NO products and cytokines havebeen studied in EBC in children with asthma. The majority ofstudies indicated that pH was lower and that eicosanoids and Th2interleukins such as IL-4 and IL-5 were elevated in children withasthma and even higher during exacerbations. However, moststudies were cross sectional and compared healthy children toasthmatic children, or children with stable asthma versus childrenwith an asthma exacerbation. IL-5 and pH were able to predictexacerbations in asthmatic children [78]. However, the applicationof EBC markers for titrating asthma treatment in individualchildren has not been studied to date.

With metabolomic approaches, detection of complete EB or EBCprofiles instead of single markers has become possible [79]. How-ever, these studies are still in an early phase for standardizationof procedures. More longitudinal studies and external validationof specific breath metabolomic profiles for certain diseases isurgently needed.

In the future, the analysis of EB and EBC might prove ofadditional value in personalized medicine but obviously it is notclear if these methods can be used in clinical practice.

E-HEALTH

E-health is a widely used term for healthcare practice supportedby electronic processes and communication and encompasses a

range of services or systems at the edge of medicine/healthcare andinformation technology. E-health has raised great expectations inmonitoring of all types of chronic diseases like diabetes,hypertension and asthma. The terms telemonitoring or telemedi-cine are frequently used interchangeably with E-health. A recentCochrane meta-analysis on effectiveness of telehealthcare inasthma identified 21 studies in adults and/or children (n = 7), allinvestigating telephone and video- and Internet-based models ofcare [80]. The authors concluded that telehealthcare did notimprove patients’ quality of life or reduce the number of asthma-related emergency department visits. Also, reminder systems wereable to increase patient adherence to treatment, but did notimprove any clinical outcomes [81].

The 7 randomised controlled trials in children showed thatweb based programs are at least non-inferior to usual care, andsome showed superiority [80]. Studies varied substantially in thepopulation included, the complexity and intensity of the interven-tion, the methods used, duration of the study and primary outcomes.

It is unclear which patients adhere most to web applications,and, what is more, might benefit most from telemedicine. Van derMeer et al studied adolescents with persistent asthma and showedthat patients with poor asthma control were more able and readyto incorporate internet-based asthma self-management for a longerperiod of time versus patients with good control [82]. Interestingly,the same group showed that involving adolescents in Internetbased self-management programs is challenging, as only 90 patientsout of almost 700 eligible patients could be randomised [83]. Therewas considerable dropout in this study, which seemed to beassociated with better asthma control.

Additionally, several randomized controlled trials aimed toimprove asthma knowledge of children and/or caregivers byinteractive educational computer programs [84–86]. In summary,these studies show improved asthma knowledge and some showimproved asthma outcomes like reduced symptoms and less ERvisits.

It might be hypothesized that Internet programs should focuson uncontrolled patients or patients with severe asthma, but manyquestions remain to be answered.

First, patients who will benefit most from e-health interven-tions will have to be identified and, once identified, ways toincrease engagement in web-based interventions should be foundwithout engendering dependence on professional/technologicalsupport. In this respect, patient preferences should be thoroughlyexamined in order to optimise web interventions [87].

Second, research should be focused on the most (cost) effectivetele-intervention taking into account at least the patient’s age,asthma severity and control state, patient needs, resources andlocal facilities.

Third, the optimal duration of web based programs, the optimalcontact frequency and feedback strategy should be determinedfor every intervention in itself and for every individual patient.Delegation of tasks to less expensive health care providers mayimprove cost-effectiveness. The interventions studied should beeasy to implement in daily care, should be reimbursed and allefforts should be made to link patient data obtained by e-healthinterventions directly to electronic patient files.

In addition, social media offer an excellent opportunity toconnect with children and adolescents; to date, no studies havebeen published on the effect of social media on asthma outcomes.

DAILY LIFE TRIALS IN PERSONALIZED MEDICINE

Although randomized controlled trials provide the highest levelof evidence, it has been suggested that only a minority of patientstreated for asthma meet the eligibility criteria for major trials[88]. Participants follow an intensive schedule of study visits,

M.W. Pijnenburg, S. Szefler / Paediatric Respiratory Reviews 16 (2015) 101–107 105

receive detailed instructions to increase adherence and aresupervised by trained study staff. Therefore, complementary waysof research like comparative effectiveness research and pragmatictrials, measuring effectiveness in routine clinical practice, maybe needed to develop a personalized approach to care [88]. Com-parative effectiveness research compares existing health careinterventions to determine which strategy works best for whichpatients under which circumstances and which pose the greatestbenefits or harms. To date, no exclusive paediatric studies haveused comparative effectiveness research to study predictors ofasthma outcomes in individual children. One study examined thereal-life effectiveness of ICS versus LTRA monotherapy in childrenwith mild or moderate asthma. In contrast to recommendationsin guidelines [3,4] it was concluded that ICS did not significantlyreduce rescue oral corticosteroids or acute care visits in compari-son with LTRA, and was associated with a higher rate of admissionsfor asthma and rescue beta-2 agonist use [89]. These findings maybe due to better adherence to treatment with LTRA. Large datasets, enhanced patient registries and access to electronic medicalrecords are essential for this type of research.

FUTURE DIRECTIONS FOR PERSONALIZED MEDICINE INCHILDREN WITH ASTHMA?

Several steps are needed to move forward with a personal-

ized asthma management approach for children.

1. We need reliable, enhanced electronic medical records and

easy access to them for physicians. [90] Decision support

tools based on current knowledge may help physicians to

decide on the best treatment for their individual patients.

2. A greater number of genes identified for each asthma drug

pathway may improve our ability to predict drug treatment

response in individual patients.

3. Better phenotyping methods for patients with asthma may

allow for stratified medicine and may help to build decision

support tools.

4. Validated and reproducible biomarkers may be helpful in

discovering asthma pathways and develop new asthma

drugs. Biomarkers are potentially helpful in selection of

drugs, monitoring treatment response, in particular to new

and expensive biologicals, and predicting events like exacer-

bations or decline in lung function.

CONCLUSIONS

To date, much has been learned about the ability to selectmedications based on the use of asthma characteristics, biomark-ers and genetics. Some of this information, for example, theapplication of sputum eosinophils and exhaled nitric oxide indecision-making, can be applied in clinical practice. However, theapplications of genetic markers and electronic health systems needadditional studies before they can be routinely applied for supportof clinical decision making for asthma care.

References

[1] Asher MI, Montefort S, Bjorksten B, Lai CK, Strachan DP, Weiland SK, et al.Worldwide time trends in the prevalence of symptoms of asthma, allergicrhinoconjunctivitis, and eczema in childhood: ISAAC Phases One and Threerepeat multicountry cross-sectional surveys. Lancet 2006;368:733–43.

[2] Bacharier LB, Boner A, Carlsen KH, Eigenmann PA, Frischer T, Gotz M, et al.Diagnosis and treatment of asthma in childhood: a PRACTALL consensusreport. Allergy 2008;63:5–34.

[3] British Thoracic Society Scottish Intercollegiate Guidelines. British Guidelineon the Management of Asthma. Thorax 2008: 63 Suppl 4: iv1–121.

[4] GINA. GINA report Global Strategy for Asthma Management and Prevention,updated 2011. 2011.

[5] National Asthma Education and Prevention Program. Expert Panel Report 3(EPR-3): Guidelines for the Diagnosis and Management of Asthma-SummaryReport 2007. J Allergy Clin Immunol 2007: 120: S94-138.

[6] Voorend-van Bergen S, Vaessen-Verberne AA, Landstra AM, Brackel HJ, van denBerg NJ, Caudri D, et al. Monitoring childhood asthma: Web-based diaries andthe asthma control test. J Allergy Clin Immunol 2014;133:1599–605. e1592.

[7] Gustafsson PM, Watson L, Davis KJ, Rabe KF. Poor asthma control in children:evidence from epidemiological surveys and implications for clinical practice.Int J Clin Pract 2006;60:321–34.

[8] Cowan K, Guilbert TW. Pediatric asthma phenotypes. Curr Opin Pediatr2012;24:344–51.

[9] Nuijsink M, De Jongste JC, Pijnenburg MW. Will symptom-based therapy beeffective for treating asthma in children? Curr Allergy Asthma Rep 2013;13:421–6.

[10] Blackwell LS, Marciel KK, Quittner AL. Utilization of patient-reported out-comes as a step towards collaborative medicine. Paediatr Respir Rev 2013;14:146–51.

[11] Klok T, Kaptein AA, Duiverman EJ, Brand PL. High inhaled corticosteroidsadherence in childhood asthma: the role of medication beliefs. Eur Respir J2012;40:1149–55.

[12] Williams B, Steven K, Sullivan FM. Tacit and transitionary: an exploration ofpatients’ and primary care health professionals’ goals in relation to asthma. SocSci Med 2011;72:1359–66.

[13] Irwin RS, Richardson ND. Patient-focused care: using the right tools. Chest2006;130:73S–82S.

[14] Hitchcock W, Mellon M, Memran M, Parasuraman B, Ramachandran S, WalzerS. Caregiver preferences for pediatric asthma treatment delivery systems. AdvTher 2007;24:1240–53.

[15] Forno E, Lescher R, Strunk R, Weiss S, Fuhlbrigge A, Celedon JC, ChildhoodAsthma Management Program Research Group. Decreased response to inhaledsteroids in overweight and obese asthmatic children. J Allergy Clin Immunol2011;127:741–9.

[16] Rogers AJ, Tantisira KG, Fuhlbrigge AL, Litonjua AA, Lasky-Su JA, Szefler SJ, et al.Predictors of poor response during asthma therapy differ with definition ofoutcome. Pharmacogenomics 2009;10:1231–42.

[17] Wu AC, Tantisira K, Li L, Schuemann B, Weiss ST, Fuhlbrigge AL, ChildhoodAsthma Management Program Research Group. Predictors of symptoms aredifferent from predictors of severe exacerbations from asthma in children.Chest 2011;140:100–7.

[18] Chauhan BF, Ducharme FM. Anti-leukotriene agents compared to inhaledcorticosteroids in the management of recurrent and/or chronic asthma inadults and children. Cochrane Database Syst Rev 2012;5. CD002314.

[19] Szefler SJ, Phillips BR, Martinez FD, Chinchilli VM, Lemanske RF, Strunk RC,et al. Characterization of within-subject responses to fluticasone and mon-telukast in childhood asthma. J Allergy Clin Immunol 2005;115:233–42.

[20] Zeiger RS, Szefler SJ, Phillips BR, Schatz M, Martinez FD, Chinchilli VM, et al.Response profiles to fluticasone and montelukast in mild-to-moderate persis-tent childhood asthma. J Allergy Clin Immunol 2006;117:45–52.

[21] Knuffman JE, Sorkness CA, Lemanske Jr RF, Mauger DT, Boehmer SJ, MartinezFD, et al. Phenotypic predictors of long-term response to inhaled corticoster-oid and leukotriene modifier therapies in pediatric asthma. J Allergy ClinImmunol 2009;123:411–6.

[22] Rabinovitch N, Graber NJ, Chinchilli VM, Sorkness CA, Zeiger RS, Strunk RC,et al. Urinary leukotriene E4/exhaled nitric oxide ratio and montelukastresponse in childhood asthma. J Allergy Clin Immunol 2010;126:545–51.e541-4.

[23] Zeiger RS, Mauger D, Bacharier LB, Guilbert TW, Martinez FD, Lemanske Jr RF,et al. Daily or intermittent budesonide in preschool children with recurrentwheezing. N Engl J Med 2011;365:1990–2001.

[24] Martinez FD, Chinchilli VM, Morgan WJ, Boehmer SJ, Lemanske Jr RF, MaugerDT, et al. Use of beclomethasone dipropionate as rescue treatment for childrenwith mild persistent asthma (TREXA): a randomised, double-blind, placebo-controlled trial. Lancet 2011;377:650–7.

[25] Papi A, Nicolini G, Baraldi E, Boner AL, Cutrera R, Rossi GA, et al. Beclometha-sone, Salbutamol Treatment for Children Study G. Regular vs prn nebulizedtreatment in wheeze preschool children. Allergy 2009;64:1463–71.

[26] Turpeinen M, Nikander K, Pelkonen AS, Syvanen P, Sorva R, Raitio H, et al. Dailyversus as-needed inhaled corticosteroid for mild persistent asthma (TheHelsinki early intervention childhood asthma study). Arch Dis Child 2008;93:654–9.

[27] Lemanske Jr RF, Mauger DT, Sorkness CA, Jackson DJ, Boehmer SJ, Martinez FD,et al. Step-up therapy for children with uncontrolled asthma receiving inhaledcorticosteroids. N Engl J Med 2010;362:975–85.

[28] Rabinovitch N, Mauger DT, Reisdorph N, Covar R, Malka J, Lemanske Jr RF, et al.Predictors of asthma control and lung function responsiveness to step 3 thera-py in children with uncontrolled asthma. J Allergy Clin Immunol 2014;133:350–6.

[29] Malka J, Mauger DT, Covar R, Rabinovitch N, Lemanske Jr RF, Spahn JD, et al.Eczema and race as combined determinants for differential response to step-up asthma therapy. J Allergy Clin Immunol 2014;134:483–5.

[30] Vaessen-Verberne AA, van den Berg NJ, van Nierop JC, Brackel HJ, Gerrits GP,Hop WC, et al. Combination therapy salmeterol/fluticasone versus doubling

M.W. Pijnenburg, S. Szefler / Paediatric Respiratory Reviews 16 (2015) 101–107106

dose of fluticasone in children with asthma. Am J Respir Crit Care Med2010;182:1221–7.

[31] Chang TS, Lemanske Jr RF, Mauger DT, Fitzpatrick AM, Sorkness CA, Szefler SJ,et al. Childhood asthma clusters and response to therapy in clinical trials. JAllergy Clin Immunol 2014;133:363–9.

[32] Zacharasiewicz A, Wilson N, Lex C, Erin EM, Li AM, Hansel T, et al. Clinical use ofnoninvasive measurements of airway inflammation in steroid reduction inchildren. Am J Respir Crit Care Med 2005;171:1077–82.

[33] Li AM, Tsang TW, Lam HS, Sung RY, Chang AB. Predictors for failed dosereduction of inhaled corticosteroids in childhood asthma. Respirology2008;13:400–7.

[34] Pijnenburg MW, Hofhuis W, Hop WC, De Jongste JC. Exhaled nitric oxidepredicts asthma relapse in children with clinical asthma remission. Thorax2005;60:215–8.

[35] Rank MA, Branda ME, McWilliams DB, Johnson SK, Samant SA, Podjasek JC,et al. Outcomes of stepping down asthma medications in a guideline-basedpediatric asthma management program. Ann Allergy Asthma Immunol2013;110:354–8. e352.

[36] Corren J, Lemanske RF, Hanania NA, Korenblat PE, Parsey MV, Arron JR, et al.Lebrikizumab treatment in adults with asthma. N Engl J Med 2011;365:1088–98.

[37] Sorkness CA, Wildfire JJ, Calatroni A, Mitchell HE, Busse WW, O’Connor GT,et al. Reassessment of omalizumab-dosing strategies and pharmacodynamicsin inner-city children and adolescents. J Allergy Clin Immunol Pract 2013;1:163–71.

[38] Lipworth BJ, Basu K, Donald HP, Tavendale R, Macgregor DF, Ogston SA, et al.Tailored second-line therapy in asthmatic children with the Arg(16) genotype.Clin Sci (Lond) 2013;124:521–8.

[39] Zuurhout MJ, Vijverberg SJ, Raaijmakers JA, Koenderman L, Postma DS, Koppel-man GH, et al. Arg16 ADRB2 genotype increases the risk of asthma exacerba-tion in children with a reported use of long-acting beta2-agonists: results ofthe PACMAN cohort. Pharmacogenomics 2013;14:1965–71.

[40] Palmer CN, Lipworth BJ, Lee S, Ismail T, Macgregor DF, Mukhopadhyay S.Arginine-16 beta2 adrenoceptor genotype predisposes to exacerbations inyoung asthmatics taking regular salmeterol. Thorax 2006;61:940–4.

[41] Basu K, Palmer CN, Tavendale R, Lipworth BJ, Mukhopadhyay S. Adrenergicbeta(2)-receptor genotype predisposes to exacerbations in steroid-treatedasthmatic patients taking frequent albuterol or salmeterol. J Allergy ClinImmunol 2009;124:1188–94. e1183.

[42] Martin AC, Zhang G, Rueter K, Khoo SK, Bizzintino J, Hayden CM, et al. Beta2-adrenoceptor polymorphisms predict response to beta2-agonists in childrenwith acute asthma. J Asthma 2008;45:383–8.

[43] Isaza C, Sepulveda-Arias JC, Agudelo BI, Arciniegas W, Henao J, Porras GL, et al.beta(2) -adrenoreceptor polymorphisms in asthmatic and non-asthmaticschoolchildren from Colombia and their relationship to treatment response.Pediatr Pulmonol 2012;47:848–55.

[44] Szczepankiewicz A, Breborowicz A, Sobkowiak P, Kramer L, Popiel A. Role ofADRB2 gene polymorphism in asthma and response to beta(2)-agonists inPolish children. J Appl Genet 2009;50:275–81.

[45] Giubergia V, Gravina L, Castanos C, Chertkoff L. Influence of beta(2)-adrenergicreceptor polymorphisms on asthma exacerbation in children with severeasthma regularly receiving salmeterol. Ann Allergy Asthma Immunol2013;110:156–60.

[46] Bonini M, Permaul P, Kulkarni T, Kazani S, Segal A, Sorkness CA, et al. Loss ofsalmeterol bronchoprotection against exercise in relation to ADRB2 Arg16Glypolymorphism and exhaled nitric oxide. Am J Respir Crit Care Med2013;188:1407–12.

[47] Moore PE, Ryckman KK, Williams SM, Patel N, Summar ML, Sheller JR. Geneticvariants of GSNOR and ADRB2 influence response to albuterol in African-American children with severe asthma. Pediatr Pulmonol 2009;44:649–54.

[48] Litonjua AA, Lasky-Su J, Schneiter K, Tantisira KG, Lazarus R, Klanderman B,et al. ARG1 is a novel bronchodilator response gene: screening and replicationin four asthma cohorts. Am J Respir Crit Care Med 2008;178:688–94.

[49] Tantisira KG, Lake S, Silverman ES, Palmer LJ, Lazarus R, Silverman EK, et al.Corticosteroid pharmacogenetics: association of sequence variants in CRHR1with improved lung function in asthmatics treated with inhaled corticoste-roids. Hum Mol Genet 2004;13:1353–9.

[50] Tantisira KG, Hwang ES, Raby BA, Silverman ES, Lake SL, Richter BG, et al.TBX21: a functional variant predicts improvement in asthma with the use ofinhaled corticosteroids. Proc Natl Acad Sci U S A 2004;101:18099–104.

[51] Tantisira KG, Lasky-Su J, Harada M, Murphy A, Litonjua AA, Himes BE, et al.Genomewide association between GLCCI1 and response to glucocorticoidtherapy in asthma. N Engl J Med 2011;365:1173–83.

[52] Koster ES, Maitland-van der Zee AH, Tavendale R, Mukhopadhyay S, VijverbergSJ, Raaijmakers JA, et al. FCER2 T2206C variant associated with chronicsymptoms and exacerbations in steroid-treated asthmatic children. Allergy2011;66:1546–52.

[53] Tantisira KG, Silverman ES, Mariani TJ, Xu J, Richter BG, Klanderman BJ, et al.FCER2: a pharmacogenetic basis for severe exacerbations in children withasthma. J Allergy Clin Immunol 2007;120:1285–91.

[54] Hawkins GA, Lazarus R, Smith RS, Tantisira KG, Meyers DA, Peters SP, et al. Theglucocorticoid receptor heterocomplex gene STIP1 is associated with im-proved lung function in asthmatic subjects treated with inhaled corticoste-roids. J Allergy Clin Immunol 2009;123:1376–83. e1377.

[55] Park HW, Dahlin A, Tse S, Duan QL, Schuemann B, Martinez FD, et al. Geneticpredictors associated with improvement of asthma symptoms in response toinhaled corticosteroids. J Allergy Clin Immunol 2014;133:664–9. e665.

[56] Mastalerz L, Nizankowska E, Sanak M, Mejza F, Pierzchalska M, Bazan-Socha S,et al. Clinical and genetic features underlying the response of patients withbronchial asthma to treatment with a leukotriene receptor antagonist. Eur JClin Invest 2002;32:949–55.

[57] Sampson AP, Siddiqui S, Buchanan D, Howarth PH, Holgate ST, Holloway JW,et al. Variant LTC(4) synthase allele modifies cysteinyl leukotriene synthesisin eosinophils and predicts clinical response to zafirlukast. Thorax 2000;55(Suppl 2):S28–31.

[58] Asano K, Shiomi T, Hasegawa N, Nakamura H, Kudo H, Matsuzaki T, et al.Leukotriene C4 synthase gene A(-444)C polymorphism and clinical responseto a CYS-LT(1) antagonist, pranlukast, in Japanese patients with moderateasthma. Pharmacogenetics 2002;12:565–70.

[59] Lima JJ, Zhang S, Grant A, Shao L, Tantisira KG, Allayee H, et al. Influence ofleukotriene pathway polymorphisms on response to montelukast in asthma.Am J Respir Crit Care Med 2006;173:379–85.

[60] Klotsman M, York TP, Pillai SG, Vargas-Irwin C, Sharma SS, van den Oord EJ, et al.Pharmacogenetics of the 5-lipoxygenase biosynthetic pathway and variableclinical response to montelukast. Pharmacogenet Genomics 2007;17:189–96.

[61] Mougey EB, Feng H, Castro M, Irvin CG, Lima JJ. Absorption of montelukast istransporter mediated: a common variant of OATP2B1 is associated withreduced plasma concentrations and poor response. Pharmacogenet Genomics2009;19:129–38.

[62] Dweik RA, Boggs PB, Erzurum SC, Irvin CG, Leigh MW, Lundberg JO, et al.American Thoracic Society Committee on Interpretation of Exhaled NitricOxide Levels for Clinical A. An official ATS clinical practice guideline: inter-pretation of exhaled nitric oxide levels (FENO) for clinical applications. Am JRespir Crit Care Med 2011;184:602–15.

[63] Baraldi E, Azzolin NM, Zanconato S, Dario C, Zacchello F. Corticosteroids decreaseexhaled nitric oxide in children with acute asthma. J Pediatr 1997;131:381–5.

[64] Artlich A, Hagenah JU, Jonas S, Ahrens P, Gortner L. Exhaled nitric oxide inchildhood asthma. Eur J Pediatr 1996;155:698–701.

[65] Byrnes CA, Dinarevic S, Shinebourne EA, Barnes PJ, Bush A. Exhaled nitric oxidemeasurementsinnormalandasthmaticchildren.Pediatr Pulmonol 1997;24:312–8.

[66] Taylor DR, Pijnenburg MW, Smith AD, De Jongste JC. Exhaled nitric oxidemeasurements: clinical application and interpretation. Thorax 2006;61:817–27.

[67] Petsky HL, Cates CJ, Lasserson TJ, Li AM, Turner C, Kynaston JA, et al. Asystematic review and meta-analysis: tailoring asthma treatment on eosino-philic markers (exhaled nitric oxide or sputum eosinophils). Thorax 2012;67:199–208.

[68] Bossley CJ, Saglani S, Kavanagh C, Payne DN, Wilson N, Tsartsali L, et al.Corticosteroid responsiveness and clinical characteristics in childhood diffi-cult asthma. Eur Respir J 2009;34:1052–9.

[69] Szefler SJ, Mitchell H, Sorkness CA, Gergen PJ, O’Connor GT, Morgan WJ, et al.Management of asthma based on exhaled nitric oxide in addition to guideline-based treatment for inner-city adolescents and young adults: a randomisedcontrolled trial. Lancet 2008;372:1065–72.

[70] Fleming L, Bush A. Use of sputum eosinophil counts to guide management inchildren with severe asthma. Thorax 2012;67:1015–6. author reply 1016.

[71] Lex C, Jenkins G, Wilson NM, Zacharasiewicz A, Erin E, Hansel TT, et al. Doessputum eosinophilia predict the response to systemic corticosteroids inchildren with difficult asthma? Pediatr Pulmonol 2007;42:298–303.

[72] Caldeira M, Perestrelo R, Barros AS, Bilelo MJ, Morete A, Camara JS, et al.Allergic asthma exhaled breath metabolome: a challenge for comprehensivetwo-dimensional gas chromatography. J Chromatogr A 2012;1254:87–97.

[73] Caldeira M, Barros AS, Bilelo MJ, Parada A, Camara JS, Rocha SM. Profilingallergic asthma volatile metabolic patterns using a headspace-solid phasemicroextraction/gas chromatography based methodology. J Chromatogr A2011;1218:3771–80.

[74] Dallinga JW, Robroeks CM, van Berkel JJ, Moonen EJ, Godschalk RW, Jobsis Q,et al. Volatile organic compounds in exhaled breath as a diagnostic tool forasthma in children. Clin Exp Allergy 2010;40:68–76.

[75] Robroeks CM, van Berkel JJ, Dallinga JW, Jobsis Q, Zimmermann LJ, Hendriks HJ,et al. Metabolomics of volatile organic compounds in cystic fibrosis patientsand controls. Pediatr Res 2010;68:75–80.

[76] Robroeks CM, van Berkel JJ, Jobsis Q, van Schooten FJ, Dallinga JW, Wouters EF,et al. Exhaled volatile organic compounds predict exacerbations of childhoodasthma in a 1-year prospective study. Eur Respir J 2013;42:98–106.

[77] Thomas PS, Lowe AJ, Samarasinghe P, Lodge CJ, Huang Y, Abramson MJ, et al.Exhaled breath condensate in pediatric asthma: promising new advance orpouring cold water on a lot of hot air?. a systematic review. Pediatr Pulmonol2013;48:419–42.

[78] Robroeks CM, van Vliet D, Jobsis Q, Braekers R, Rijkers GT, Wodzig WK, et al.Prediction of asthma exacerbations in children: results of a one-year prospec-tive study. Clin Exp Allergy 2012;42:792–8.

[79] Carraro S, Rezzi S, Reniero F, Heberger K, Giordano G, Zanconato S, et al.Metabolomics applied to exhaled breath condensate in childhood asthma. AmJ Respir Crit Care Med 2007;175:986–90.

[80] McLean S, Chandler D, Nurmatov U, Liu J, Pagliari C, Car J. Telehealthcare forasthma: a Cochrane review. CMAJ 2011;183:E733–42.

[81] Tran N, Coffman JM, Sumino K, Cabana MD. Patient reminder systems and asthmamedication adherence: a systematic review. J Asthma 2014;51:536–43.

[82] van der Meer V, van Stel HF, Detmar SB, Otten W, Sterk PJ, Sont JK. Internet-based self-management offers an opportunity to achieve better asthma con-trol in adolescents. Chest 2007;132:112–9.

[83] Rikkers-Mutsaerts ER, Winters AE, Bakker MJ, van Stel HF, van der Meer V, deJongste JC, et al. Internet-based self-management compared with usual care in

M.W. Pijnenburg, S. Szefler / Paediatric Respiratory Reviews 16 (2015) 101–107 107

adolescents with asthma: a randomized controlled trial. Pediatr Pulmonol2012;47:1170–9.

[84] Homer C, Susskind O, Alpert HR, Owusu M, Schneider L, Rappaport LA, et al. Anevaluation of an innovative multimedia educational software program forasthma management: report of a randomized, controlled trial. Pediatrics2000;106:210–5.

[85] Krishna S, Francisco BD, Balas EA, Konig P, Graff GR, Madsen RW. Internet-enabled interactive multimedia asthma education program: a randomizedtrial. Pediatrics 2003;111:503–10.

[86] Sockrider MM, Abramson S, Brooks E, Caviness AC, Pilney S, Koerner C, et al.Delivering tailored asthma family education in a pediatric emergency depart-ment setting: a pilot study. Pediatrics 2006;117:S135–44.

[87] Baptist AP, Thompson M, Grossman KS, Mohammed L, Sy A, Sanders GM. Socialmedia, text messaging, and email-preferences of asthma patients between12 and 40 years old. J Asthma 2011;48:824–30.

[88] Price D, Hillyer EV, van der Molen T. Efficacy versus effectiveness trials:informing guidelines for asthma management. Curr Opin Allergy Clin Immunol2013;13:50–7.

[89] Ducharme FM, Noya FJ, Allen-Ramey FC, Maiese EM, Gingras J, Blais L. Clinicaleffectiveness of inhaled corticosteroids versus montelukast in children withasthma: prescription patterns and patient adherence as key factors. Curr MedRes Opin 2012;28:111–9.

[90] Weiss ST. New approaches to personalized medicine for asthma: where arewe? J Allergy Clin Immunol 2012;129:327–34.

Paediatric Respiratory Reviews 16 (2015) 108–109

Cochrane Corner

Pancreatic enzyme replacement therapy for people withcystic fibrosis (Review)

U.R. Somaraju 1,*, A. Solis-Moya 2

1 Department of Biochemistry, Malla Reddy Medical College for Women, Hyderabad, India2 Servicio de Neumologıa, Hospital Nacional de Ninos, San Jose, Costa Rica

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

WHY IT IS IMPORTANT TO DO THE REVIEW

Pancreatic enzyme insufficiency is seen in approximately 80%to 90% of people with cystic fibrosis (CF) and results in maldigestionand malnutrition [1]. Malnutrition is associated with poor growthand development in children, poor weight gain in adults, moresevere pulmonary disease and shorter life expectancy [2,3]. Peoplewith CF have been treated with pancreatic enzyme replacementtherapy (PERT) since the 1930s with the aim of improving nutritionalstatus and clinical outcome [4]. Currently there are several PERTpreparations available including non-enteric-coated powders, en-teric-coated tablets (ECT), enteric-coated microspheres (ECM),enteric-coated minimicrospheres, and enteric coated micro tablets[5]. They differ in composition, formulation, and bioavailability. Therequired daily dose of PERT is highly variable and depends on thepatient’s age, body weight, genotype and degree of pancreaticinsufficiency as well as the type and quantity of meals [4]. The efficacyof PERT also depends on the coating and size of enzyme particles inpreparations, gastric transit time, gastric and small intestinal pH, andlack of bile acids [6]. Current dosing recommendations for PERT areintended to prevent harm, rather than treat effectively leading towide variation in clinical practice [7]. Preparations are generallyconsidered safe, but are associated with potential significant sideeffects. It is therefore very important to establish the evidence forthe benefits and risks of PERT; to compare different formulations;to determine the optimum treatment for different age groups; andto clarify the role of tests of pancreatic function in therapy [8].

WHAT COMPARISONS DID WE MAKE IN THE REVIEW?

We compared PERT (started at any time after diagnosis) given inany dose or formulation, for at least four weeks to placebo oranother PERT preparation.

* Corresponding author.

http://dx.doi.org/10.1016/j.prrv.2014.11.001

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

We identified 18 potentially eligible randomized controlledtrials and included 12 of these in the review [9–20]. The trialsincluded a total of 426 participants. Two authors independentlyassessed the trials for inclusion in the review and contacted thetrial authors for further information.

We reported data by comparison of interventions e.g. ECM toother enteric-coated and non-enteric-coated preparations, anddifferent doses of PERT (Table 1). We planned to perform asubgroup analysis looking at the different formulations anddosages, the presence of symptoms and if possible look at effectsat different ages. However, this was not possible due to the smallnumber of included trials.

RESULTS

Eight trials included children with CF and their ages ranged from1 to 17 years [9–16]. The participants in the remaining four trialswere adults with mean ages varying between 21.4 and 24.8 years[17–20].

The interventions were heterogeneous between the trials;10 trials compared ECM with other preparations of PERT, includingother ECM. Two trials compared PERT in different doses (Table 1).We could only combine data from six trials in the analysis, as theother six trials did not have sufficient information. The risk of biasfor the methods of the included studies was classified largely as‘‘unclear’’ since the description of their methods was inadequate toassess the risk of bias; we judged half of the trials to be at ‘‘highrisk’’ of bias with respect to selective reporting of results.

Primary outcomes of interest for the review were absolute orrelative change in nutritional status (weight, height, BMI). Therewas a small increase in weight in people taking ECM whencompared to non-enteric-coated tablets.

Secondary outcomes included bowel symptoms (stool frequen-cy, abdominal pain, flatulence, constipation, distal intestinalobstruction syndrome), days in hospital, quality of life, number

Table 1Number of studies included in the review using different kinds and doses of PERT preparations.

Comparisons No.of studies Reference

ECM vs non-enteric coated preparations Lyophilized Total Pancreatic Extracts 1 Vidailhit 1987 [14]

Non-enteric coated tablets 2 Henker 1987 [10], Stead 1987 [20]

ECM vs other enteric coated preparations ECT 2 Stead 1986 [19], Vyas 1990 [15]

Enteric coated granules 1 Peterson 1984 [13]

ECM 3 Elliott 1992 [9], Williams 1990 [16], Lacy 1992 [11]

Enteric coated minimicrospheres 1 Patchell 1999 [12]

Different doses of PERT High dose vs low dose 1 Assoufi 1994 [17]

Three doses 1 Borowitz 2005 [18]

Figure 1. Forest plot of comparison: ECM versus ECT, outcome: FFE [g/day].

U.R. Somaraju, A. Solis-Moya / Paediatric Respiratory Reviews 16 (2015) 108–109 109

of diagnoses of vitamin deficiency, adverse events attributed toPERT, fecal fat excretion (FFE) or co-efficient of fat absorption(Figure 1) and lung disease. The people with CF taking ECMexperienced decreases in stool frequency, in abdominal pain and inFFE, when compared to non-enteric coated preparations and ECT.

There was no significant difference in any of the outcomes fromthe different preparations of ECM or enteric-coated minimicro-spheres.

IMPLICATIONS FOR PRACTICE

The available evidence suggests that ECM are better atimproving clinical symptoms in patients with CF compared tonon-enteric-coated enzyme preparations. This evidence is, how-ever, limited and is from a few small trials that are prone to bias.

Further research is needed to study the different forms of PERT,their efficacy, safety, and role in improving nutritional status,quality of life and their long-term effects. In addition it is alsonecessary to investigate if the same amount of PERT is applicable toall ranges of pancreatic insufficiencies in CF.

In collaboration with the Cochrane CF and Genetic Disorders Group’

http://cfgd.cochrane.org/.

References

[1] Dodge JA, Turck D. Cystic fibrosis: nutritional consequences and management.Best practice & research. Clinical Gastroenterology 2006;20(3):531–46.

[2] Corey M, McLaughlin FJ, Williams M, Levinson H. A comparison of the survival,growth and pulmonary function in patients with cystic fibrosis. Journal ofClinical Epidemiology 1988;41:583–91.

[3] Stallings V, Stark L, Robinson K, Feranchak A, Quinton H. Evidence-basedpractice recommendations for nutrition-related management of childrenand adults with cystic fibrosis and pancreatic insufficiency: results of asystematic review. Journal of the American Dietetic Association 2008;108(5):832–9.

[4] Imrie CW, Connett G, Hall RJ, Charnley RM. Review article: enzyme supple-mentation in cystic fibrosis, chronic pancreatitis, pancreatic and periampul-lary cancer. Alimentary Pharmacology & Therapeutics 2010;32(Suppl. 1):1–25.

[5] Fieker A, Philpott J, Armand M. Enzyme replacement therapy for pancreaticinsufficiency: present and future. Clinical and Experimental Gastroenterology2011;4:55–73. http://dx.doi.org/10.2147/CEG.S17634.

[6] Li Li, Somerset S. Digestive system dysfunction in cystic fibrosis: Challenges fornutrition therapy. Digestive and Liver Disease 46 (2014) 865–874.

[7] Haupt ME, Kwasny MJ, Scheschter MS, McColley SA. Pancreatic EnzymeReplacement Therapy Dosing and Nutritional Outcomes in Children withCystic Fibrosis. J Pediatr 2014;164:1110–5.

[8] Somaraju UR, Solis-Moya A. Pancreatic enzyme replacement therapy forpeople with cystic fibrosis. Cochrane Database of Systematic Reviews 2014,Issue 10. Art. No.: CD008227. DOI:10.1002/14651858.CD008227.pub2.

[9] Elliott RB, Escobar LC, Lees HR, Akroyd RM, Reilly HC. A comparison of twopancreatin microsphere preparations in cystic fibrosis. New Zealand MedicalJournal 1992;105(930):107–8.

[10] Henker J, Jutta Hein L, Vogt E. Comparison of effectiveness of Pankreon ForteRand KreonR in children with cystic fibrosis [abstract]. 15th Annual Meeting ofthe European Working Group for Cystic Fibrosis; 1987. 1987.p. 72.

[11] Lacy D, West J, Venkataraman M, Vyas J, MacDonald A, Weller PH, et al. Acomparison of Nutrizym GR and Creon in children with CF [abstract]. 11thInternational Cystic Fibrosis Congress; 1992. 1992. TP76.

[12] Patchell CJ, Desai M, Weller PH, MacDonald A, Smyth RL, Bush A, et al. Creon110000 minimicrospheresTM vs. Creon1 8000 microspheres - an open ran-domized crossover preference study. Journal of Cystic Fibrosis 2002;1(4):287–91.

[13] Petersen W, Heilmann C, Garne S. Pancreatic enzyme supplementation as acid-resistant microspheres versus enteric-coated granules in cystic fibrosis. Adouble placebo- controlled crossover study. Acta Paediatrica Scandinavica1987;76(1):66–9.

[14] Vidailhet M, Derelle J, Morali A, De Gasperi JP. Comparison of effectiveness andtolerance of enteric coated versus unprotected pancreatic extracts in CFpatients [abstract]. 15th Annual Meeting of the European Working Group forCystic Fibrosis; 1987. 1987.p. 69.

[15] Vyas H, Matthew DJ, Milla PJ. A comparison of enteric coated microsphereswith enteric-coated tablet pancreatic enzyme preparations in cysticfibrosis. A controlled study. European Journal of Pediatrics 1990;149(4):241–3.

[16] Williams J, MacDonald A, Weller PH, Fields J, Pandov H. Two enteric coatedmicrospheres in cystic fibrosis. Archives of Disease in Childhood 1990;65(6):594–7.

[17] Assoufi BK, Doig C, Hodson ME. High dose Nutrizym 22 in adults with cysticfibrosis [abstract]. Pediatric Pulmonology 1994;18(Suppl 10):337.

[18] Borowitz D, Goss CH, Limauro S, Konstan MW, Blake K, Casey S, et al.Study of a novel pancreatic enzyme replacement therapy in pancreaticinsufficient subjects with cystic fibrosis. Journal of Pediatrics 2006;149(5):658–62.

[19] Stead RJ, Skypala I, Hodson ME, Batten JC. Enteric coated microspheres ofpancreatin in the treatment of cystic fibrosis: comparison with a standardenteric coated preparation. Thorax 1987;42(7):533–7.

[20] Stead RJ, Skypala I, Hodson ME. Treatment of steatorrhoea in cystic fibrosis: acomparison of enteric-coated microspheres of pancreatin versus non-enteric-coated pancreatin and adjuvant cimetidine. Alimentary Pharmacology & Ther-apeutics 1988;2(6):471–82.

In collaboration with the Cochrane CF and Genetic Disorders Group’ http://cfgd.cochrane.org/.

Paediatric Respiratory Reviews 16 (2015) 110–111

Cystic Fibrosis Frequently Asked Questions

Question 1: Does antibiotic susceptibility testing matter when choosing anantibiotic for a pulmonary exacerbation?

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

Antibiotic susceptibility testing (AST) has been the basis forselecting antibiotic treatment of pulmonary infections andexacerbations in individuals with cystic fibrosis [1]. However,there is emerging evidence suggesting that testing the antibioticsusceptibility of CF pathogens against various antibiotics may notbe predictive of a clinical response.

A retrospective review of patients attending the CF unit inNottingham Children’s Hospital found no correlation betweenantibiotic concordance based on individual AST and changes inlung function, nutritional status, or time to next exacerbation[1]. However, the reality is children with a pulmonary exacerbationwho have previously isolated Pseudomonas aeruginosa willusually concurrently receive two anti-pseudomonal antibioticsas the mainstay of treatment.

Thus, the more pertinent question is whether combinationantibiotic susceptibility testing (CAST) contributes to clinicaloutcome. A randomized, double-blind, controlled clinical trialassessing the effects of antibiotic therapy directed by combinationantibiotic susceptibility testing was conducted [2]. Treatmentsbased on CAST did not improve clinical or bacteriologicaloutcomes, and were no more effective than treatment based onconventional culture and sensitivity testing [2]. As such, adoptionof CAST as a routine practice for CF patients infected with multi-resistant pathogens is viewed as costly and unnecessary [3].

The limited clinical utility of antibiotic susceptibility testingmay be due to the fact that microbiological testing routinelyinvolves the assessment of planktonically–cultured organismsobtained from CF patients, and thus, results of antibioticsusceptibility testing may not reflect the actual susceptibilitiesof organisms that exist in the lung in biofilms [4]. Hence,therapeutic antibiotics chosen based on the tests of antibioticsusceptibility of planktonic bacteria instead of biofilm –existingbacteria may incorrectly predict in vivo response to antibiotictreatment [5,6].

Efforts have been made to consider antibiotic regimens basedon biofilm susceptibility. A study conducted by Aaron et al. [5]examining the single and combination antibiotic susceptibilities ofPseudomonas aeruginosa in differing modes of growth found thatthe Pseudomonas aeruginosa in a biofilm was more resistant to thebactericidal effects of both single and multiple combinations ofantibiotics than the planktonically grown Pseudomonas aerugi-nosa isolates.

Similarly, Dales et al. [6] examined the antibiotic susceptibili-ties of Burkholderia cepacia and Pseudomonas. aeruginosa grownwith biofilms to both double and triple antibiotic combinations.

http://dx.doi.org/10.1016/j.prrv.2015.01.004

1526-0542/

They found that biofilm – grown bacterial isolates were moreresistant to the same antibiotic combinations used on panktoni-cally – grown bacterial isolates. Triple antibiotic combinationswere also found to be more effective than double antibioticcombinations against B. cepacia and P. aeruginosa isolates withinbiofilms. In addition to bacterial states, Hill et al. [7] found that thenumber of antibiotic combinations that demonstrated bactericidaleffects differed depending upon laboratory culture conditions. Thisstudy illustrated that there were significantly more antibioticbactericidal combinations for isolates grown under aerobicconditions, than for isolates grown under anaerobic conditions.Both anaerobically grown organisms and organisms grown withina biofilm were significantly more resistant to single agents andantibiotic combinations than corresponding aerobic planktonicallygrown isolates. This further supported the view that antibioticsusceptibility patterns achieved under planktonic conditions invitro were not necessarily bactericidal against bacteria grown inbiofilm - enriched conditions. The use of biofilm susceptibilitytesting may more accurately reflect the bacterial growth condi-tions in the CF airways, and selection of antibiotic combinationsbased on results of biofilm susceptibility testing may be moreeffective in reducing the bacterial load in the airways of a CFpatient.

Investigating whether there is a strong relationship between invitro biofilm antibiotic susceptibility results and in vivo clinicalefficacy outcomes has also been undertaken [8]. This retrospectivestudy assessed whether CF patients infected with multi-resistantpathogens treated with antibiotics that inhibited the biofilm-grown pathogens found indirect evidence suggesting that patientstreated with biofilm-effective therapy appeared to have improvedclinical outcomes [8]. A significant decrease in sputum bacterialdensity and duration of hospital stay was observed in the group ofpatients treated with biofilm-susceptible antibiotics. Lung func-tion (FEV1 and FVC) and dyspnoea scores after 14 days of antibiotictreatment were found to have improved, as well as time to the nextpulmonary exacerbation, but these results lacked statisticalsignificance, possibly due to the small sample size [4]. Additionally,the retrospective nature of this study would suggest thatprospective studies should be undertaken to further investigatethese encouraging early observations.

On a more speculative note, there is evidence that the CF lungbacterial microbiome may have an impact on antibiotic resistance[9,10]. Altering the microbiota has been associated with infectionand potentially resistance to anti-microbial treatment. This mayexplain the discordance between sputum culture species isolation,

Cystic Fibrosis Frequently Asked Questions / Paediatric Respiratory Reviews 16 (2015) 110–111 111

antibiotic sensitivity results and the patient’s response toantimicrobial treatment [11]. Indeed, as suggested by Dickson[9], the impact of the acute and chronic use of antibiotic therapiesmay not so much ‘‘clear’’ an infection but ‘‘alter the internalarchitecture of a dynamic and heterogeneous microbial commu-nity’’. It has been demonstrated that with age and CF diseaseseverity as reflected by progression of airway obstruction, bacterialdiversity decreases, especially in people with homozygousp.F508.del CF [12]. A recent study by Zhao et al (2012) highlightedthat recent antibiotic use was the best predictor of reduceddiversity in the airway microbiome [13]. There is clearly a need tourgently broaden our knowledge of the influence of the airway andgut microbiome on CF airway infection and the impact of antibiotictreatment on microbial diversity.

References

[1] Hurley MN, Amin Ariff AH, Bertenshaw C, Bhatt J, Smyth AR. Results ofantibiotic susceptibility testing do not influence clinical outcome in childrenwith cystic fibrosis. Journal of Cystic Fibrosis 2012;11:288–92.

[2] Aaron SD, Vandemheen KL, Ferris W, Fergusson D, Tullis E, Haase D, et al.Combination antibiotic susceptibility testing to treat exacerbations of cysticfibrosis associated with multiresistant bacteria: a randomised, double-blinded, controlled clinical trial. The Lancet 2005;366:463–71.

[3] Aaron SD. Antibiotic synergy testing should not be routine for patients withcystic fibrosis who are infected with multiresistant bacterial organisms.Paediatric Respiratory Review 2007;8:256–61.

[4] Caraher E, Reynolds G, Murphy P, McClean S, Callaghan M. Comparison ofantibiotic susceptibility of Burkholderia cepacia complex organisms whengrown planktonically or as biofilm in vitro. European journal of clinical micro-biology & infectious diseases: official publication of the European Society of ClinicalMicrobiology 2006;26:213–6.

[5] Aaron S, Ferris W, Ramotar K, Vandemheen K, Chan F, Saginur R. Single andCombination antibiotic susceptibilities of planktonic adherent, and bio-filmgrown pseudomonas aeruginosa isolates cultured from sputa of adults withcystic fibrosis. J Clin Microbiol 2002;40:4172–9.

[6] Dales L, Ferris W, Vandemheen K, Aaron S. Combination antibiotic suscepti-bility of biofilm-grown Burkholderia cepacia and Pseudomonas aeruginosaisolated from patients with pulmonary exacerbations of cystic fibrosis.

European journal of clinical microbiology & infectious diseases: official publicationof the European Society of Clinical Microbiology 2009;28:1275–9.

[7] Hill D, Rose B, Pajkos A, Robinson M, Bye P, Bell S, et al. Antibiotic susceptibili-ties of Pseudomonas aeruginosa isolates derived from patients with cysticfibrosis under aerobic, anaerobic, and biofilm conditions. J Clin Microbiol2005;43:5085–90.

[8] Keays T, Ferris W, Vandemheen KL, Chan F, Yeung SW, Mah TF, et al. Aretrospective analysis of biofilm antibiotic susceptibility testing: A betterpredictor of clinical response in cystic fibrosis exacerbations. Journal of CysticFibrosis 2009;8:122–7.

[9] Dickson RP, Erb-Downward JR, Huffnagle GB. The role of the Bacterial Micro-biome in Lung Disease. Expert Rev Respir Med 2013;7:245–57. http://dx.doi.org/10.1586/ers.13.24.

[10] Twomey KB, Alston M, An S-Q, O’Connell OJ, McCarthy Y, et al. Microbiotiaand Metabollite Profiling Reveal Specific Alterations in Bacterial CommunityStructure and Environment in the Cystic Fibrois Aiarway during Exacerbation.PLoS ONE 2013;8(12):e82432. http://dx.doi.org/10.1371/journal.pone.0082432.

[11] Smith AL, Fiel SB, Mayer-Hamblett N, Ramsey B, Burns JL. Susceptibility testingof Pseudomonas aeruginosa isolates and clinical response to parenteral anti-biotic administration: Lack of association in cystic fibrosis. Chets 2003;123:1495–502.

[12] Cox MJ, Allgaier M, Taylor B, et al. Airway microbiota and pathogen abundancein age-stratified cystic fibrosis patients. PLoS ONE 2010;5(6):e11044.

[13] Zhao J, Schloss PD, Kalikin LM, et al. Decade-long bacterial communitydynamics in cystic fibrosis airways. Proceedings of the National Academy ofScience 2012;109(15):5809–14.

Sandy Z.P. LimMedical Student, Sydney Medical School, University of Sydney

Dominic A. Fitzgerald1,2,*1The Children’s Hospital at Westmead, Sydney

2Sydney Medical School, Discipline Paediatrics & Child Health,

University of Sydney

*Corresponding author. Department of Respiratory Medicine,The Children’s Hospital at Westmead, Locked Bag 4001,

Westmead, NSW, 2145.Tel.: +61 2 9845 3397; fax: +61 2 9845 3396

E-mail address: Dominic.fitzgerald@health.nsw.gov.au

Paediatric Respiratory Reviews 16 (2015) 112–118

Review

Pertussis in the Newborn: certainties and uncertainties in 2014

Gustavo Rocha *, Paulo Soares, Henrique Soares, Susana Pissarra, Hercılia Guimaraes

Division of Neonatology, Department of Pediatrics, Hospital de Sao Joao, Porto, Portugal

A R T I C L E I N F O

Keywords:

Antibiotic resistance

Bordetella pertussis

epidemiology

hyperleukocytosis

laboratory diagnosis

newborn

pulmonary hypertension

vaccination.

S U M M A R Y

Bordetella pertussis infection remains a serious potential health risk to infants, specially in those too

young to be vaccinated. Over the recent years, numerous sources highlighted a widespread resurgence,

making it, again, a challenging disease. Globally, pertussis is ranked among the 10 leading causes of

childhood mortality. This review summarizes the most recent literature and will address the most

important aspects that pediatricians and neonatologists must be familiar with, when treating a newborns

pertussis infection.

� 2014 Elsevier Ltd. All rights reserved.

EDUCATIONAL AIMS

� This review summarizes the most recent literature and addresses the most important aspects that pediatricians and neonatologistsmust be familiar with, when treating a newborn’ pertussis infection.� Clinical aspects, laboratory diagnosis, treatment, epidemiology over the last years, and vaccination aspects are highlighted.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

INTRODUCTION

Pertussis, or whooping cough, is caused by the organismBordetella pertussis and remains a serious potential health risk toinfants, specially in those too young to be vaccinated [1]. Althoughremoved from the list of notifiable diseases in many countries onceapparentely conquered after routine vaccination, over recentyears, numerous sources highlighted a widespread resurgencemaking it, again, a challenging disease [2,3].

Severe respiratory failure complicated by pulmonary hyperten-sion is associated with high mortality which may develop in thenewborn and young children [4]. Globally, pertussis is rankedamong the 10 leading causes of childhood mortality [5]. Macrolideresistance has been documented and is an essencial point wheninvestigating individual treatment failures [6]. New vaccinationstrategies have been suggested in the last decade, predominantlyto protect infants younger than two months of age [7].

This review summarizes the most recent literature and willaddress the most important aspects that pediatricians and

* Corresponding author. Servico de Neonatologia/Departamento de Pediatria,

Hospital de Sao Joao – Piso 2, Alameda Prof. Hernani Monteiro, 4200 – 319 Porto,

Portugal Tel.: +351 225512100x1949; fax: +351225512273.

E-mail address: gusrocha@oninet.pt (G. Rocha).

1526-0542/$ – see front matter � 2014 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.prrv.2014.01.004

neonatologists must be familiar with when treating a pertussis

infection in a young infant.

PATHOPHYSIOLOGY, CLINICAL PRESENTATION ANDDIFFERENTIAL DIAGNOSIS

Bordetella pertussis is a small aerobic Gram-negative coccoba-cillus that infects only humans. It causes irritation and inflamma-tion by infecting the ciliated respiratory tract epithelium. Thebacterium has several virulence factors and toxins that areimportant in the pathogenesis of the disease and also play a rolein inducing protective response. Filamentous hemagglutinin andfimbriae are adhesins required for tracheal colonization. Theensuing tissue necrosis and epithelial cell damage recruitsmacrophages, and reactive lymphoid hyperplasia of peribronchialand tracheobronchial lymph nodes occur. Other virulence factorssuch as pertactin and pertussis toxin can act as adhesins as well [8].

Pertussis toxin can inactivate or supress signaling pathways ofthe immune system in the lung, which delays recruitment ofneutrophils. The role of pertussis toxin in the pathogenesis ofpertussis is not fully understood. The toxin has been shown to causeleukocytosis with lymphocytosis and possibly the rare encephalo-pathy seen in clinical disease. Other direct effects of pertussis toxininclude sensitization of the beta-islet cells of the pancreas. Thiseffect can lead to hyperinsulinemia with a resistant hypoglycemia,

G. Rocha et al. / Paediatric Respiratory Reviews 16 (2015) 112–118 113

and sometimes occurs in young infants who have poor feeding.Adenylate cyclase toxin inhibits migration and activation ofphagocytes and T cells [8].

Pertussis is spread by aerossol droplets expelled whilecoughing or sneezing. Infants get pertussis from older siblings,parents, or caregivers who may have only mild symptoms [9].After an incubation period of seven to 14 days, the natural historyof pertussis tends to follow a relatively predictable clinical course,although disease severity and prognosis are quite variable, and ahigh degree of suspicion may be necessary to make a timelydiagnosis. A child suspected of having pertussis should be placedin appropriate isolation until the infection is confirmed or ruledout [8].

The catarrhal phase lasts one to two weeks, and includesnonspecific complaints. The absent or mild fever, cough and nasalsigns are similar to those seen in many viral upper respiratory tractinfections, which often delays the suspicion of the diagnosis. Thecough worsens as the patient progresses to the paroxysmal phase.This phase of illness lasts from weeks two to six, and ischaracterized by paroxysms of typical cough. The paroxysmsmay occur several times per hour and can be associated withcyanosis, salivation, lacrimation, and post-tussive emesis. They canbe exhausting and often interfere with sleep and nutritional intake.Newborns and young infants often have a less typical presentation.The classic ‘‘whoop’’ following the paroxysm of cough may beabsent, and gasping, gagging, and apnea can occur. The clinicalpicture of the most severely affected newborns and young infantsmay be dominated by marked respiratory distress, cyanosis, andapnea, rather than significant cough. Lymphocyte counts arefrequently in excess of 30 x109/L [10]. Leukocytosis, together withan absolute lymphocytosis on a peripheral complete blood count,is a laboratory finding suppotive of Bordetella pertussis infection.This finding often correlates with disease severity, specially in veryyoung patients. White blood cell counts as high as 30 to 60 x109/Lcan be seen [8].

The convalescent phase follows the paroxysmal phase withimprovement in respiratory tract integrity and function, decreas-ing frequency and severity of the coughing episodes, and may lastfrom weeks to months [8].

Complications of pertussis include apnea, pneumonia, pulmon-ary hypertension, seizures, encephalopathy, pneumothorax, pneu-momediastinum, subcutaneous emphysema, rib fracture, emesisand dehydration, hypoglycaemia, rectal prolapse, superficialpetechial hemorrhage, and even intracranial hemorrhage. Bacterialand viral superinfection such as influenza or respiratory syncytialvirus may occur, and can lead to a more severe clinical course[8,10]. Bordetella pertussis pneumonia may progress rapidly andpulmonary hypertension may result in right-sided heart failure orfatal cardiac arrythmias [10]. Malignant pertussis is defined by arapidly evolving combination of pneumonia, respiratory failure,severe leukocytosis, neurologic envolvement, and finally, severepulmonary hypertension leading to death in 75% of cases, despiteintensive therapeutic measures. The mortality rate of all infantpertussis infection is about 1%, with more than half of the deathsoccurring in infants below two months age [11].

Respiratory tissue samples obtained at autopsy from 15 infantsaged below four months who had polymerase chain reaction- orculture-confirmed pertussis pneumonia were evaluated by multi-ple histochemical stains, immunohistochemical evaluation, andelectron microscopic examination, by Paddock CD et al [4]. Thepulmonary histopathologic examination of the samples revealed adescending infection dominated by necrotizing bronchiolitis,intra-alveolar hemorrhage, and fibrinous edema. All sampleshad marked leukocytosis and most showed luminal aggregates ofabundant leukocytes in small pulmonary arteries, veins, andlymphatics. A novel immunohistochemical stain for Bordetella

pertussis revealed abundant extracellular Bordetellae in cilia of thetrachea, bronchi, and bronchioles, as well as intracellular bacteriaand antigens in alveolar macrophages and ciliated epithelium.

Encephalopathy associated with pertussis infection is rare,occurring in 0.5–1% of all cases, and the diagnoses is suggested byseizures with pertussis infection, in the absence of other diagnosis.It may be the result of the direct neurologic actions of toxins,effects of hypoxia, hemorrhages, vascular occlusions and latentvirus infection. However, the aetiology of central nervous systemcomplications are not fully understood [12,13]. The syndrome ofinnapropriate secretion of antidiuretc hormone (SIADH) includinghyponatremia and decreased urine output, although rare, hasalready been decribed in association with pertussis infection[14,15].

The differential diagnosis includes infections from otherrespiratory pathogens, as Respiratory Syncitial Virus, influenza,adenoviruses, and Chlamidia trachomatis. Rapid viral antigentesting and specific serologies will help these diagnosis [8].

PERTUSSIS-RELATED HYPOXEMIA, HYPERLEUKOCYTOSIS, ANDPULMONARY HYPERTENSION

Bordetella pertussis infection may develop a fulminant course invery young, unimmunized infants, and is characterized bypneumonia that rapidly evolves to respiratory failure withrefractory hypoxemia, extreme leukocytosis and cardiogenic shockrequiring cardiovascular support [4,8,15–19]. The main risk factorsfor high mortality are high white blood cell (WBC) count andsevere pulmonary hypertension (PH), age below six months,prematurity, and incomplete immunization [17].

The setting up of refractory hypoxemia characteristically israpid and responds inadequately to advanced ventilation man-euvers, including high-frequency oscillatory ventilation, inhalednitric oxide, and extracorporeal membrane oxygenation (ECMO)[18,19]. There are some hypotheses for this poor outcome. Anassociation between high WBC count and increase in severity led tothe theory that vascular infiltration or hyperviscosity may be afactor in inducing PH and heart failure [16,20].

The mechanism for the induction of intractable PH by pertussis

infection is unknown. Hypoxia contributes to PH, but other factorssuch as toxin production or pulmonary venous leukocyte thrombidemand additional investigation [21]. Though, the lack of PH insome cases lead to the theory that primary ventricular dysfunctionmay be a separate contributory entity and could also be the focus oftherapeutic intervention [16].

Histological examination of the lungs of fatal cases of pertussis

infection exhibited histopathologic features of necrotizing bronch-iolitis, bronchopneumonia, pulmonary hemorrhage and edema,angiolymphatic leukocytosis, and revealed numerous bacteria inintracellular and extracellular compartments of airways andairspaces. The association between vascular obstruction (bloodviscosity, and the consequent tendency for microthrombusformation) and bronchial obstruction (mucus plugs) could beresponsible for hypoxemic respiratory failure, acidosis, and PHcausing right ventricular failure [16,22]. Some post-mortem studiesin patients aged below four months found pulmonary leukocytethrombi obstructing the lumen of the arteries, veins andlymphatics [23]. In the youngest infants with reactive pulmonaryarterioles and immature coagulation and fibrinolytic systems,severe leukocytosis contributes directly to severity of diseasethrough a hyperviscosity syndrome and pulmonary arteriolarthrombosis. This microvascular thrombotic obstruction to pul-monary blood flow is resistant to conventional management ofpulmonary hypertension with pulmonary vasodilators; in fact,inhaled nitric oxide may worsen the effects of the pertussis trachealcytotoxin [24]. Founded on the assumption that the pulmonary

G. Rocha et al. / Paediatric Respiratory Reviews 16 (2015) 112–118114

vascular leukostasis was responsible for PH and respiratoryfailure, some authors advocate the use of leukoreductiontechniques [23].

Increased WBC count is observed in virtually all cases ofpertussis. Hyperleukocytosis (>100x109/L) is an unusual compli-cation and may be attributed to pertussis toxins [17]. Theseproteins act together to stimulate lymphocyte proliferation andactivation and to elevate cAMP levels which may contribute topulmonary vasoconstriction [20]. Another possible reason forlymphocytosis is the decrease in the expression of L-selectin byleukocytes in infants with pertussis infection. The L-selectinexpression is important in the extravasation of leukocytes intotissues and for homing peripheral blood lymphocytes to lymphnodes [17].

The pertussis has been associated with the degree of leukocy-tosis, whereby a strikingly elevated WBC count is related tocardiopulmonary compromise. This warrants the use of intensivemeans in order to reduce the excessive leukocytes. Some authorsconsider the use of hyperhydration for all hyperleukocytosis casesand suggest the use of exchange transfusion when WBC count isover 100x109/L [17,19,24]. Lowering WBC counts by leukofiltra-tion or exchange transfusion provides some evidence that astrategy of leukodepletion may be associated with improvedsurvival [19,24,25]. A double-volume blood exchange transfusionor leukopheresis has allowed successful treatment of malignantpertussis in infants by decreasing leukocyte mass and platelets, aswell as other possible deleterious circulating inflammatorymediators, the decrease in blood viscosity, and the consequentreduction in the tendency for microthrombus formation and forthe adhesion of activated leukocytes to the vascular endotheliumwith subsequent vascular obstruction [22].

The use of ECMO with a leukocyte filter or double-exchangetransfusion for infants with extreme leukocytosis has beenreported. In a series of 10 patients, the use of these techniqueswas associated with a greater-than-expected survival ratecompared with historical controls, but studies are lacking torecommend its regular clinical use [26].

Infants and children in respiratory failure referred for ECMOtherapy often have systemic hypotension, probably resulting fromhypoxia induced myocardial depression, given that it generallyimproves as systemic oxygenation get better. The causes ofsystemic hypotension may also include the release of systemicvasodilators, like in the systemic inflammatory response syndrome[27].

The ECMO has been used in children for the treatment of severerespiratory failure and PH secondary to infection by Bordetella

pertussis, with a mortality rate of around 70% in treated cases[15,26,28,29]. Infants who are younger than six weeks andhave pertussis have the worst outcomes (83.6% case fatality rate)[25,31]. When comparing mortality rates in patients withpertussis with those on ECMO for other reasons, the observedmortality with pertussis was higher. Indeed, overall ECMOmortality rates varied from 14% to 35% for respiratory causes[21,24].

Risk factors for pertussis mortality include ventilator strategyand acidosis before the start of ECMO. Survivors had receivedhigher PEEP than nonsurvivors, probably reflecting better ventila-tion in infants who ultimately survived. Pre-ECMO pH was lower ininfants who died [21,30].

Although ECMO has proven to be valuable in treating persistentPH, the PH associated with pertussis has not always responded toECMO and the explanation for this still remains unclear [28]. Forthis reason, probably, the mechanisms leading to increasedpulmonary vascular resistance in many patients with severepertussis are different from those associated with persistent PH ofthe newborn [27].

In summary, ECMO and exchange transfusion should bediscussed in the setting of clinical deterioration, increased whiteblood cells and PH secondary to severe pertussis in infants [30].Although ECMO should still be offered as a therapeutic modality,physicians should discuss with parents that the majority of theseinfants do not survive [21,30].

LABORATORY DIAGNOSIS

Laboratory diagnosis of Bordetella pertussis infection is challen-ging, in part due to the wide diversity of methods used worldwide[31,32]. It relies on direct and indirect tests. The former (nucleicacid amplification tests – namely polymerase chain reaction (PCR)– and culture) aiming at detecting microorganism presencewhereas indirect tests (serology) detect the presence of antibodiesagainst Bordetella pertussis antigens, therefore diagnosing past andpresent infection as well as previous vaccination.

CULTURE

Bordetella pertussis culture remains the gold standard forlaboratory diagnosis of infection due to its high specificity(100%). However, its low sensitivity, mainly because the Bordetella

is a highly labile microorganism that often does not survive sampletransport and processing. Sampling for culture is difficult andinfluences sensitivity as does the timing during the course ofdisease at which sampling occurs (highest if samples are obtainedat the catarrhal phase of the disease). The best samples for cultureresult from nasopharyngeal aspirates (since Bordetella pertussis hasa special affinity for the ciliated respiratory epithelium) immedi-ately cultured onto adequate culture medium (charcoal agarenriched in horse blood and cephalexin 40 mg/L and Bordet-Gengou medium). If a nasopharyngeal swab is to be obtained, aDacron swab should be preferred, but this yields a lowersensitivity. The use of transport medium also lowers culturesensitivity but is often necessary due to difficulties with bedsideculture. Another drawback of culture is the time required for apositive result. Plates should be incubated for seven days and thereis usually visible growth after three days. Bordetella pertussis is acatalase positive, oxidase positive and urease negative gramnegative cocco baccilus [33]. Sensitivity is also influenced bypatient age (highest for young unvaccinated infants) and symptomduration (highest for shorter duration of symptoms).

DIRECT ANTIGEN DETECTION – IMMUNOFLUORESCENCE

Although a more rapid tool for the diagnosis of Bordetella

pertussis infection, immunofluorescence is generally not acceptedas proof of infection due to its lack of sensitivity and specificity[33].

SEROLOGY

Recommendations concerning serological diagnosis of Borde-

tella pertussis have recently been revised and published by aEuropean Panel of Experts – the EU Pertstrain Group [34].Detection of Bordetella pertussis serum antibodies is the basis ofserologic diagnosis. Antibodies can be detected in blood, serumand oral fluids. Most commercially available assays are validated totest serum, since the diagnostic sensitivity of measuring antibodiesin oral fluid samples is lower (about 80%) than that of serumsamples [35]. EU Pertstrain Group recommends that, if possible,acute and convalescent serum samples taken at least three weeksapart should be tested together in one run in order to allow a moreclinically meaningful interpretation of results. Enzyme-linkedimmunosorbent assay (ELISA) has been the gold standard for

G. Rocha et al. / Paediatric Respiratory Reviews 16 (2015) 112–118 115

serological diagnosis of Bordetella pertussis infection. There areseveral Bordetella pertussis antibodies that can be detected bycurrently available assays, although only pertussis toxin antibodiesare specific and unique for Bordetella pertussis. ELISA should bedone with purified non –detoxified pertussis (PT) toxin antigens,but other antigens, like filamentous haemagglutinin (FHA),pertractin (PRN) and fimbriae (FIM) are occasionally also used.These other antigens are less specific for Bordetella pertussis sincethey cross react with other microbial antigens (other Bordetella

species, Haemophilus influenza, Chlamydia pneumoniae, Mycoplasm

pneumoniae and Escherichia coli) and therefore the use of mixedantigen ELISA kits for the diagnosis of Bordetella pertussis infectionas well as the use of immunoblots, complement fixation, indirectimmunofluorescence, microagglutination assays and cell neutra-lization assays should be strongly discouraged [34,36]. Naturalinfection with Bordetella pertussis induces the production of IgA,IgG and IgM antibodies whereas immunization is followed by anincrease in the serum concentration of IgG and IgM antibodies [37].The interpretation of serological results is hampered by the factthat immune response against infection or vaccination may beindistinguishable since most acellular pertussis vaccines containBordetella pertussis antigens PT, FHA, PRN and FIM, thus inducingan immunological response characterized by the serum elevationof anti- PT antibodies. Dual sample serology showing more than100% increase or more than a 50% decrease in antibodyconcentration reveals a sensitive and specific method to diagnoseinfection, although, in clinical practice, diagnosis is usually basedon single sample serology, which is more sensitive and morespecific than paired serology [38]. Results of measurement ofantibodies to Bordetella pertussis antigens should be reported in IU/ml, according to World Health Organization International Standard(06/140) and a cut-off for IgG anti PT between 50 and 120 IU/ml (atleast one year after vaccination with acellular pertussis vaccines)should be considered diagnostic of Bordetella pertussis infection[39]. Higher cut-offs, while improving specificity of the test, losesensitivity and are less appropriate for diagnosis. IgA antibodiesmeasurement seems appealing when the diagnosis of infectioncannot be ascertained based on a single serum IgG anti-PTdetermination and a second convalescent serum is not available.However, no broadly accepted cut-offs exist for this antibodysubclass which is highly specific but with very poor sensitivity andthus the lower limit of quantitation for the anti-PT IgA ELISA (12IU/ml) has often been used [40,41]. An important drawback in theinterpretation of Bordetella pertussis antibody detection pattern isthe age dependent antibody isotype response following vaccina-tion. Unlike individuals that experience a first contact withBordetella pertussis through vaccination, which produce only IgGisotype anti PT antibody, vaccination of individuals who havepreviously been exposed to Bordetella pertussis induces theproduction of both IgG and IgA antibodies, which makes theantibody reactivity pattern of these vaccinated individualsindistinguishable from that of acute infection [42]. This aspect isof less importance in newborns and young children, usuallyunvaccinated prior to exposure to Bordetella pertussis. Despitethese drawbacks, serology, single or dual, is the most efficientdiagnostic test, with a relatively high sensitivity and a highspecificity in children and adults but in neonates and young infantsits use is hampered by time constraints and immunologicalimmaturity.

NUCLEIC ACID AMPLIFICATION TESTS

Nucleic acid amplification tests (NAAT) such as polymerasechain reaction (PCR) and loop-mediated isothermal amplification(LAMP) have become routine methods of Bordetella pertussis

infection diagnosis, replacing culture and direct fluorescent

antibody staining of nasopharyngeal secretions due to theirincreased sensitivity [43]. Nasopharyngeal aspirates are thepreferred specimens for NAAT, since they allow recovery of higheramounts of microorganisms than swabs. PCR tests for the detectionof Bordetella were first described in 1989 [44] and soon thereafterthey were shown to be more sensitive than culture, the goldstandard [45]. The use of loop mediated isothermal amplificationmethods to detect Bordetella pertussis has been described. Thismethod was 100 times more sensitive than PCR and highly specificfor the detection of Bordetella pertussis and not other Bordetella

species, thus presenting an advantage over other tests [46]. Since itdoes not require specific equipment, this method promises tobecome a useful tool for the rapid diagnosis of Bordetella pertussis

infection in the clinical setting [47]. NAAT are the most rapid andsensitive laboratory methods for the diagnosis of Bordetella

pertussis infection. Its sensitivity, however, decreases substanciallywith increasing patient age and duration of symptoms[48]. Whenconsidering NAAT for the detection of Bordetella pertussis, genetargets species specificity are very important. Most PCR tests arebased on the detection of insertion sequences (IS) that are presentin variable number in the genome of different Bordetella species. Ofthese, the most widely used is IS481, present in more than 50copies in the Bordetella pertussis genome [49]. This IS is, however,also present in Bordetella holmesii genome in 8 to 10 copies pergenome, and in animal isolates (and, even less frequently, humanisolates derived from immunocompromised individuals) of Borde-

tella bronchiseptica making it necessary to exclude these agents inthe presence of an IS481 positive result. This issue has beenrecently addressed in several studies aiming at the development ofmultitarget assays that will allow a clear identification of Bordetella

species [50]. Other insertion sequences- IS 1001, hIS1001 e IS1002-allow the diagnosis of infection with Bordetella species other thanBordetella pertussis [51]. Single gene targets, namely the promoterregion of pertussis toxin operon, pertactin gene sequences,filamentous hemagglutinin gene, adenylate cyclase gene, outermembrane porin gene, Rec A protein gene, flagelin gene, BP3385gene, thiolase gene BP283, BP485 gene, can also be tested, butpresent lower sensitivity than multi-copy targets and, with theexception of the promoter region of pertussis toxin operon, assaysbased on single gene targets have not been extensively evaluatedand are not routinely used in clinical practice [52]. The pertussis

toxin operon is also present though not expressed in otherBordetella species (Bordetella parapertussis and bronchiseptica) andthus a positive PCR result requires a post-amplification analysis inorder to exclude the presence of these other species. On thecontrary, the pertussis toxin operon is not present in Bordetella

holmesii which allows the distinction of Bordetella pertussis andBordetella holmesii using a combination of primers targeting thepertussis toxin operon and IS481 (only Bordetella pertussis beingpositive for both) [53]. Unlike pertussis toxin operon, the pertussis

toxin promoter (ptxA-Pr) is a target for Bordetella pertussis specificassays, although consistently less sensitive than IS481. Bordetella

pertussis DNA remains detectable for a variable period of time ininfants less than six months of age and real-time PCR may be usefulto diagnose pertussis infection up to three weeks after treatmentinitiation [54,55].

ANTIBIOTIC TREATMENT

Antibiotic treatment should be reserved for patients withBordetella pertussis identified from culture or PCR. Erythromycin (a14-day course of 40-50 mg/kg/d divided in four doses), is thetreatment of choice of Bordetella pertussis infection. However, ininfants aged below one month, the use of erythromycin isassociated with the occurrence of pyloric hypertrophic stenosis.For that reason, in infants aged below one month, erythromycin

G. Rocha et al. / Paediatric Respiratory Reviews 16 (2015) 112–118116

and clarithromycin are not recommended and azithromycin is tobe preferred (oral administration at a dose of 10 mg/kg/d oncedaily for five days). If erythromycin is to be used, infants should bemonitored for possible pyloric hypertrophic stenosis. The combi-nation trimethoprim-sulphamethoxazol is contra-indicated ininfants below a month of age. Side effects of azithromycin includegastrointestinal symptoms (abdominal discomfort and pain,diarrhea, nausea, vomiting), headache and dizziness. Cautionshould be taken in patients with impaired liver function, inpatients concomitantly using agents metabolized by the cyto-chrome P450 enzyme system. Concomitant use of aluminum andmagnesium containing antiacids should be avoided, since thesereduce the rate of azithromycin absortion [56,57]. Antibiotictreatment leads to negative cultures in five to seven days but, whenstarted after the paroxysms are established, has no discernibleeffect on the course of the illness. Although rare, macrolideresistant Bordetella pertussis has been previously described andshould be monitored [6,58]. Although the mechanism of resistanceis largely unknown, mutations of the erythromycin binding site inthe 23S rRNA gene seem to be involved [6,59].

Postexposure antimicrobial prophylaxis should be offered toasymptomatic contacts, regardless of their immunization status,within 21 days of the onset of cough in the index case in order toprevent the development of symptoms and diminish the risk oftransmission. Exposed healthcare workers and patients shouldalso receive chemoprophylaxis. Erythromycin at full dose is thedrug of choice for chemoprophylaxis, except for neonates, inwhom chemoprophylaxis should be done with azithromycin[56,57].

EPIDEMIOLOGY: CHANGES IN THE LAST YEARS

In the world, it is estimated that 50 million cases and 300.000deaths occur due to pertussis infection annually. Pertussismorbidity and mortality is most significant in infants under threemonths of age, particularly in the newborn [8].

Massive pertussis vaccination has altered the epidemiology ofthe disease. Adolescent and adult pertussis is now becomingincreasingly more frequent in countries with high vaccinationcoverage [60]. Several not mutually exclusive reasons may explainthis change, including Bordetella pertussis strain adaptation [61]and waning immunity post immunization [62]. In the prevaccina-tion era, virtually every child experienced Bordetella pertussis

infection, and because of continuous bacterial circulation, frequentbooster exposures resulted in lasting immunity against disease.Neonates may therefore have been partly protected by maternalantibodies [63].

Pertussis in adolescents and adults may represent an importantreservoir for transmission to unvaccinated infants [64]. Teenagersbooster immunization, although certainly of benefit for theteenage population, may have little impact on protection againstinfant pertussis. At the current stage of knowledge, it cannot beexcluded that vaccinating teenagers further erodes the immunityin adults of child-bearing age and thereby increases the circulationof the disease in this age group, which is the main source of severepertussis in infants [63]. Another approach would be to start DTaPimmunization at a younger age, with shorter intervals betweendoses. This schedule could be started at birth, and the first threedoses could be completed by three months of age. Notably, duringthe period of greatest reduction in pertussis incidence in the UnitedStates (1954–1974), the three-dose primary series was completedbetween three and five months of age [65].

In 2005 and 2010, substantial epidemics occurred and anotheroutbreak is now under way [65,66].

In 2008, the Collaborative Pediatric Critical Care Research Network

Critical Pertussis Study was started in the United States in order to

provide an updated overview of critical pertussis to the pediatriccritical care community. This collaborative group reported in theirfirst publication, in 2011, that despite high coverage for childhoodvaccination, pertussis causes substantial morbidity and mortalityin US children, especially among infants. In pediatric intensive careunits, Bordetella pertussis is a community-acquired pathogenassociated with critical illness and death. The incidence of medicaland developmental sequelae in critical pertussis survivors remainsunknown, and the appropriate strategies for treatment andsupport remain unclear. The Collaborative Pediatric Critical Care

Research Network Critical Pertussis Study concluded that research isurgently needed to provide an evidence base that might optimizemanagement for critical pertussis, a serious, disabling, and toooften fatal illness for U.S. children and those in the developingworld [67]. In 2013, after a prospective study including 127patients, this group concluded, in their second publication [68]that pulmonary hypertension may be associated with mortality inpertussis critical illness, elevated WBC is associated with the needfor mechanical ventilation, pulmonary hypertension, and mortal-ity risk, and research is indicated to elucidate how pulmonaryhypertension, immune responsiveness, and elevated WBC con-tribute to morbidity and mortality and whether leukoreductionmight be efficacious.

Now, in 2014, it is time to recognize the successes of the pastand to implement new directions for the control of pertussis.

VACCINATION

Routine vaccination of children and adolescents is the mostimportant preventive strategy. The national recommendationsvary considerably, and different schedules are used for the primaryseries, including immunization at the ages of six, ten and fourteenweeks; at two, three and four months; at three, four and fivemonths; and at two, four and six months. A number of countriesadminister the vaccine at three, five and twelve months: the dosesat three and five months may be considered the primary series andthe dose at twelve months a booster [69].

The morbidity and mortality in last years from pertussis imposesnew prevention strategies.These new strategies include vaccina-tion of adolescents, adults and postpartum women. Vaccination ofpregnant women and newborns is still controversial.

MATERNAL VACCINATION

The main objective of pertussis vaccination should be to protectinfants, particularly less than three months of age. The maternalantibodies transfer protection to the newborns, so it is intuitivelyreasonable to propose immunization during pregnancy [70].Administration of Pw or Pa vaccines late in pregnancy is safe forthe mother and the infant and it results in higher levels ofBordetella pertussis antibodies in infants, that may be high enoughto protect the infants through the highest risk period. However,placental transferred maternal antibodies are unlikely to persist athigh enough concentrations to sustain protection [63]. Measurableincreases in agglutinin antibodies have been found in 80–100% ofwomen immunized with Pw vaccine during pregnancy [71].Limited data suggest that Pa vaccine given to pregnant women willresult in significantly increased antibody concentrations in new-borns, but the duration of the maternal antibodies and thepotential requirement for booster doses with subsequent preg-nancies has not been sufficiently explored [69]. The CDC’s AdvisoryCommittee on Immunization Practices recommends a single doseof TdPa for pregnant women after 20 weeks gestation, if notpreviously vaccinated and if not administered during pregnancy,TdaP should be administered immediately postpartum.

G. Rocha et al. / Paediatric Respiratory Reviews 16 (2015) 112–118 117

POSTPARTUM WOMEN VACCINATION (‘‘A COCOON STRATEGY’’)

A ‘‘cocoon’’ strategy has been proposed to prevent pertussis innewborns in some countries (i.e. France, USA and Australia) [69].This strategy consists of immunizing the parents and otherhousehold contacts immediately after the birth [63]. A successfulcocoon programme implies high numbers of people to bevaccinated. In low incidence countries, cost effectiveness rate ishigh and is difficult to implement [72].

NEONATAL VACCINATION

Neonatal vaccination appears as an attractive option. Admin-istration of a three component Pa vaccine given to infants at birthand at three, five and eleven months induced a poor antibodyresponse after the first dose and efficient priming [63]. A recentstudy showed significantly lower antibody responses to pertussis

antigens upon immunization at birth with a DTPa vaccine,followed by a dose at two, four, six and 17 months, comparedto those observed in infants who were not immunized at birth [73].Early Pa vaccination may interfere with antibody responses toother vaccines (i.e. hepatitis B) [74]. The CDC’s Advisory Committeeon Immunization Practices do not recommended early neonatalimmunization.

ADOLESCENTS AND ADULTS VACCINATION

Some countries – including Australia, Canada, France, Germanyand the United States – offer adolescents and adults a singlebooster of aP vaccine in combination with tetanus toxoid andreduced-dose diphtheria vaccine, but there is no evidence thatthese programmes have had an impact on severe pertussis ininfants [69]. The CDC’s Advisory Committee on ImmunizationPractices recommends routine immunization of adolescents at 11to 12 years of age with a single dose of TdaP and a single dose ofTdaP instead of Td for adults.

VACCINATION OF HEALTH-CARE WORKERS

Several studies have shown that health-care workers are at anincreased risk of pertussis, and that transmission in health-caresettings poses substantial risk for severe nosocomial disease,particularly to infants and people who are immunocompromised.In any country where adult vaccination against pertussis isrecommended, immunization of health-care workers is logical,especially for those who have contact with unimmunized infants[69].

CONFLICTS OF INTEREST STATMENT

The authors declare that there are no conflicts of interest toreport.

References

[1] Vitek CR, Pascual FB, Baughman AL, Murphy TV. Increase in deaths frompertussis among young infants in the United States in the 1990s. Pediatr InfectDis J 2003;22:628–34.

[2] Mattoo S, Cherry JD. Molecular pathogenesis, epidemiology, and clinicalmanifestations of respiratory infections due to Bordetella pertussis and otherBordetella subspecies. Clin Microbiol Rev 2005;18:326–82.

[3] Guiso N, Liese J, Plotkin S. The global pertussis initiative: meeting report fromthe fourth regional roundtable meeting, France, April 14-15, 2010. Humanvaccin 2011;7:481–8.

[4] Paddock CD, Sanden GN, Cherry JD, Gal AA, Langston C, Tatti KM, Wu KH,Goldsmith CS, Greer PW, Montague JL, Eliason MT, Holman RC, Guarner J, ShiehWJ, Zaki SR. Pathology and Pathogenesis of fatal Bordetella pertussis infection ininfants. Clin Infect Dis 2008;47:328–38.

[5] World Health Organization. Pertussis: immunization surveillance, assessmentand monitoring.

[6] Guillot S, Descours G, Gillet Y, Etienne J, Floret D, Guiso N. Macrolide-ResistantBordetella pertussis infection in newborn girl, France. Emerg Infect Dis2012;18:966–8.

[7] Ulloa-Gutierrez R. Pertussis vaccination in newborns. Expert Rev Vaccines2009;8:153–7.

[8] Snyder J, Fisher D. Pertussis in childhood. Pediatrics in Review 2012;33:412–22.

[9] Deen JL, Mink CM, Cherry JD, Christenson PD, Pineda EF, Lewis K, Blumberg DA,Ross LA. Household contact study of Bordetella pertussis infections. Clin InfectDis 1995;21:1211–9.

[10] Barnett ED, Klein Jo. Bacterial infections of the respiratory tract. In: RemingtonJS, Klein JO, Wilson CB, Baker CJ, editors. Infectious diseases of the fetus andnewborn infant. 6th edition, Philadelphia: Elsevier Saunders; 2006. p. p299–301.

[11] Haberling DL, Holman RC, Oaddock CD, Murphy TV. Infant and maternal riskfactors for pertussis-related infant mortality in the United States, 1999 to2004. Pediatr Infect Dis J 2009;28:194–8.

[12] Hiraiwa-Sofue A, Ito Y, Mori H, Ichiyama T, Okumura A. Pertussis-associatedencephalitis/encephalopathy with marked demyelination in an unimmunizedchild. J Neurol Sci 2012;320:145–8.

[13] Chin LK, Burgner D, Buttery J, Bryant PA. Pertussis encephalopathy in an infant.Arch Dis Child 2012 Dec 18 [Epub ahead of print].

[14] Matherne P, Matson J, Marks MI. Pertussis complicated by the syndrome ofinappropriate antidiuretic hormone secretion. Pathophysiology and manage-ment. Clin Pediatr (Phila) 1986;25:46–8.

[15] Pilorget H, Montbrun A, Attali T, Tiran-Rajaofera I, Bony C, Brayer C, SamperizS, Alessandri JL. Malignant pertussis in the young infant. Arch Pediatr2003;10:787–90.

[16] Sawal M, Cohen M, Irazuzta JE, Kumar R, Kirton C, Brundler MA, Evans CA,Wilson JA, Raffeeq P, Azaz A, Rotta AT, Vora A, Vohra A, Abboud P, Mirkin LD,Cooper M, Dishop MK, Graf JM, Petros A, Klonin H. Fulminant Pertussis: AMulti-Center Study With New Insights Into the Clinico-Pathological Mechan-isms. Pediatr Pulmonol 2009;44:970–80.

[17] Lashkari HP, Karuppaswamy S, Khalifa K. Pertussis-Related Hyperleukocyto-sis: Role of Hyperhydration and Exchange Transfusion. Clin Pediatr2012;51:987–90.

[18] Romano MJ, Weber MD, Weisse ME, Siu BL. Pertussis Pneumonia, Hypoxemia,Hyperleukocytosis, and Pulmonary Hypertension: Improvement in Oxygena-tion After a Double Volume Exchange Transfusion. Pediatrics 2004;114:e264–6.

[19] Onoro G, Salido AG, Martınez IM, Cabeza B, Gillen M, de Azagra AM. Leukor-eduction in patients with severe pertussis with hyperleukocytosis. PediatrInfect Dis J 2012;31:873–6.

[20] Pierce C, Klein N, Peters M. Is leukocytosis a predictor of mortality in severepertussis infection? Intensive Care Med 2000;26:1512–4.

[21] Halasa NB, Barr FE, Johnson JE, Edwards KM. Fatal Pulmonary HypertensionAssociated With Pertussis in Infants: Does Extracorporeal Membrane Oxyge-nation Have a Role? Pediatrics 2003;112(6Pt1):1274–8.

[22] Martinez M, Rochat I, Corbelli R, Tissieres P, Rimensberger PC, Barazzone-Argiroffo C. Early blood exchange transfusion in malignant pertussis: A casereport. Pediatr Crit Care Med 2011;12:e107–9.

[23] Menif K, Bouziri A, Khaldi A. Coqueluche et hypertension arterielle pulmonairefatale. Archives de Pediatrie 2010;17:1550–2.

[24] Rowlands HE, Goldman AP, Harrington K, Karimova A, Brierley J, Cross N,Skellett S, Peters MJ. Impact of Leukodepletion on the Outcome of SevereClinical Pertussis in Young Infants. Pediatrics 2010;126:e816–27.

[25] Taffarel P, Bonetto G, Haimovich A. Coqueluche grave, evolucion y exsangui-notransfusion como tratamiento alternativo. Serie de casos. Arch Arg Pediatr2012;110:327–30.

[26] Yeh S. Treatment and prevention of Bordetella pertussis infection in infantsand children. Available at www.uptodate.com in 26-01-2013.

[27] Pooboni S, Roberts N, Westrope C, et al. Extracorporeal Life Support inPertussis. Pediatr Pulmonol 2003;36:310–5.

[28] Berry BB, Lynch JE, Chung DH. Zwischenberger. Pertussis with severe pul-monary hypertension and leukocytosis treated with extracorporeal mem-brane oxygenation. Pediatr Surg Int 2005;21:692–4.

[29] Couchot E, Paut O, Ghez O, Charpentier A, Ughetto F. Extracorporeal mem-branous oxygenation in severe infant pertussis: A case report. Ann Fr AnesthRanim 2009;28:74–7.

[30] Berthomieu L, Boumahni B, Bey KJ. La coqueluche maligne: a propos de 3observations. Archives de Pediatrie 2010;17:144–8.

[31] He Q, Barkoff AM, Mertsola J, Glismann S, Bacci S. High heterogeneity inmethods used for laboratory confirmation of pertussis diagnosis amongEuropean countries, 2010: integration of epidemiological and laboratorysurveillance must include standardization of methodologies and qualityassurance. Euro Surveill 2012;17. pii20239.

[32] Todella ML, Carlone GM, Messonier N, Quinn CP, Meade BD, Burns DL,Cherry JD, Guiso N, Hewlett EL, Edwards KM, Xing D, Giammanco A, Wirsingvon Konig CH, Han L, Hueston L, Robbins JB, Powel M, Mink CM, Poolman JT,Hildreth SW, Lynn F, Morris A. International Bordetella pertussis assaystandardization and harmonization meeting report. CDC. Vaccine 2009;27:803–14.

[33] Hallander H. Microbiological and serological diagnosis of Pertussis. Clin InfectDis 1999;28(suppl 2):S99–106.

G. Rocha et al. / Paediatric Respiratory Reviews 16 (2015) 112–118118

[34] Guiso N, Berbers G, Fry NK, He Q, Riffelmann M, Wirsing von Konig CH. What todo and not to do in serological diagnosis of pertussis: recommendations fromEU reference laboratories. Eur J Clin Microbiol Infect Dis 2011;30:307–12.

[35] Litt DJ, Samuel D, Duncan J, Harnden A, George RC, Harrison TG. Detection ofanti pertussis toxin IgG in oral fluids for use in diagnosis and surveillance ofBordetella pertussis infection in children and yong adults. J Med Microbiol2006;55(Pt9):1223–8.

[36] Xing D, Markley K, Newland P, Rigsby P, Hockley J, He Q. EUVAC.NET colla-borative study: evaluation and standardization of serology for diagnosis ofpertussis. J Immunol Methods 2011;372:137–45.

[37] Fry NK, Tzivra O, Li YT, McNiff A, Doshi N, Maple PAC, Crowcroft NS, Miller E,George RC, Harrison TG. Laboratory diagnosis of pertussis infections: the role ofPCR and serology. J Clin Microbiol 2004;53:519–25.

[38] Andre P, Caro V, Njamkepo E, Wendelboe AM, Van Rie A, Guiso N. Comparisonof serological and Real-time PCR assays to diagnose Bordetella pertussisinfection in 2007. J Clin Microbiol 2008;46:1672–7.

[39] Xing D, Wirsing von Konig CH, Newland P, Riffelmann M, Meade BD, Corbel M,Gaines-Das R. Characterization of reference materials for human antiserum topertussis antigens by an international collaborative study. Clin Vaccine Immu-nol 2009;16:303–11.

[40] Ward JI, Cherry JD, Chang SJ, Partridge S, Keitel W, Edwards K, Lee M, Treanor J,Greenberg DP, Barenkamp S, Bernstein DI, Edelman R, APERT Study Group.Bordetella pertussis infections in vaccinated and unvaccinated adolescents andadults, as assessed in a national prospective randomized Acellular PERtussisvaccine Trial (APERT). Clin Infect Dis 2006;43:151–7.

[41] Riffelmann M, Thiel K, Schmetz J, Wirsing von Konig CH. Performance ofcommercial Enzyme-Linked Immunosorbent Assays for detection of antibo-dies to Bordetella pertussis. J Clin Microbiol 2010;48:4459–63.

[42] Prince H, Lieberman JM, Cherry JD. Age-related differences in patterns ofincreased Bordetella pertussis antibodies. Clin Vaccine Immunol 2012;19:545–50.

[43] Riffelmann M, Wirsing von Konig CH, Caro V, Guiso N. Nucleic acid amplifica-tion tests for diagnosis of Bordetella infections. J Clin Microbiol 2005;43:4925–9.

[44] Houard S, Hackel C, Herzog A, Bollen A. Specific identification of Bordetellapertussis by the polymerase chain reaction. Res Microbiol 1989;140:477–87.

[45] Loeffelholz MJ, Thompson CJ, Long KS, Gilchrist MJ. Comparison of PCR, cultureand direct fluorescent antibody testing for detection of Bordetella pertussis. JClin Microbiol 1999;37:2872–6.

[46] Kamachi K, Toyoizumi-Ajisaka H, Toda K, Soeung SC, Sarath S, Nareth Y,Horiuchi Y, Kojima K, Takahashi M, Arakawa Y. Development and evaluationof a loop-mediated isothermal amplification method for rapid diagnosis ofBordetella pertussis infection. J Clin Microbiol 2006;44:1899–902.

[47] Nakamura A, Sakano T, Nakayama T, Shimoda H, Okada Y, Hanayama R,Nomoto K, Suto T, Kinoshita Y, Furue T, Ono H, Ohta T. Neonatal pertussispresenting as acute bronchiolitis: direct detection of Bordetella pertussisgenome using loop-mediated isothermal amplification. Eur J Pediatr2009;168:347–9.

[48] van der Zee A, Agterberg C, Peeters M, Mooi F, Schellekens J. A clinicalvalidation of Bordetella pertussis and Bordetella parapertussis polymerase chainreaction: comparison with culture and serology using samples from patientswith suspected whooping cough from a highly immunized population. J InfectDis 1996;174:89–96.

[49] Reischl U, Lehn N, Sanden GN, Loeffelholz MJ. Real time PCR assay targetingIS481 of Bordetella pertussis and molecular basis for detection of Bordetellaholmesii. J Clin Microbiol 2001;39:1963–6.

[50] Tatti KM, Sparks KN, Boney KO, Tondella ML. Novel multitarget real-time PCRassay for rapid detection of Bordetella species in clinical specimens. J ClinMicrobiol 2011;49:4059–66.

[51] Tizolova A, Guiso N, Guillot S. Insertion sequences shared by Bordetella speciesand implications for the biological diagnosis of pertussis syndrome. Eur J ClinMicrobiol Infect Dis 2013;32:89–96.

[52] - Loeffelholz M. Towards improved accuracy of Bordetella pertussis nucleic acidamplification tests. J Clin Microbiol 2012;50:2186–90.

[53] Tatti KM, Wu KH, Tondella ML, Cassiday PK, Cortese MM, Wilkins PP, SandenGN. Development and evaluation of dual-target real time polymerase chainreaction to detect Bordetella spp. Diagn Microbiol Infect Dis 2008;61:264–72.

[54] Bidet P, Liguori S, De Lauzanne A, Caro V, Lorrot M, Carol A, Faye A, Guiso N,Bingen E, Bonacorsi S. Real time PCR measurement of persistence of Bordetellapertussis DNA in nasopharyngeal secretions during antibiotic treatment ofyoung children with pertussis. J Clin Microbiol 2008;46:3636–8.

[55] Bonacorsi S, Farnoux C, Bidet P, Caro V, Aizenfisz S, Benhayoun M, Aujard Y,Guiso N, Bingen E. Treatment failure of nosocomial pertussis infection in avery-low-birth-weight neonate. J Clin Microbiol 2006;44:3830–2.

[56] Murphy TV, Slade BA, Broder KR, Kretsinger K, Tiwari T, Joyce Mp, Iskander JK,Brown K, Moran JS. Prevention of pertussis, tetanus and diphtheria amongpregnant and postpartum women and their infants: recommendations of theAdvisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep2008;57:1–47.

[57] Tiwari T, Murphy T, Moran J. Recommended antimicrobial agents for thetreatment and postexposure prophylaxis of pertussis: 2005 CDC Guidelines.MMWR Recomm Rep 2005;54:1–16.

[58] Fry N, Duncan G, Vaghji L, George R, Harrison T. Antimicrobial susceptibilitytesting of historical and recent clinical isolates of Bordetella pertussis in theUnited Kingdom using the Etest method. Eur Clin Microbiol Infect Dis2010;29:1183–5.

[59] Barktus J, Juni B, Ehresmann K, Miller C, Sanden G, Cassiday P, Saubolle M, LeeB, Long J, Harrison A, Besser J. Identification of a mutation associated witherythromycin resistance in Bordetella pertussis: implications for surveillance ofantimicrobial resistance. J Clin Microbiol 2003;41:1167–72.

[60] von Konig CH, Halperin SA, Riffelmann M, Guiso N. Pertussis of adults andinfants. Lancet Infect Dis 2002;2:744–50.

[61] Mooi FR. Bordetella pertussis and vaccination: the persistence of a geneticallymonomorphic pathogen. Infect Gent Evol 2010;66:594–602.

[62] Wendelboe AM, Van Rie A, Salmaso S, Englung JA. Duration of immunityagainst after natural infection or vaccination. Pediatr Infect Dis J 2005;24(5suppl):S58–61.

[63] Locht C, Mielvarek N. New pertussis vaccination approaches: en route toprotect newborns? Immunol Med Microbiol 2012;66:121–33.

[64] Lavine J, Broutin H, Harvill ET, Bjornstad ON. Imperfect vaccine-inducedimmunity and whooping cough transmission to infants. Vaccine2010;29:11–6.

[65] Cherry JD. The present and the future control of pertussis. Clin Infect Dis2010;51:663–7.

[66] Cherry JD. Epidemic pertussis in 2012-the resurgence of a vaccine-preventabledisease. N Engl J Med 2012;367–9.

[67] Burr JS, Jenkins TL, Harrison R, Meert K, Anand KJ, Berger JT, Zimmerman J,Carcillo J, Dean JM, Newth CJ, Willson DF, Sanders Jr RC, Pollack MM, Harvill E,Nicholson CE. Eunice Kennedy Shriver National Institute of Child Health andHuman Development (NICHD) Collaborative Pediatric Critical Care ResearchNetwork (CPCCRN). The Collaborative Pediatric Critical Care Research Net-work Critical Pertussis Study: collaborative research in pediatric critical caremedicine. Pediatr Crit Care Med 2011;12:387–92.

[68] Berger JT, Carcillo JA, Shanley TP, Wessel DL, Clark A, Holubkov R, Meert KL,Newth CJ, Berg RA, Heidemann S, Harrison R, Pollack M, Dalton H, Harvill E,Karanikas A, Liu T, Burr JS, Doctor A, Dean JM, Jenkins TL, Nicholson CE. EuniceKennedy Shriver National Institute of Child Health and Human Development(NICHD) Collaborative Pediatric Critical Care Research Network (CPCCRN).Critical pertussis illness in children: a multicenter prospective cohort study.Pediatr Crit Care Med 2013;14:356–65.

[69] Pertussis vaccines: WHO position paper. Weekly epidemiological record, No.40, 1ST October 2010.

[70] Mooi FR, De Greeff SC. The case for maternal vaccination against pertussis.Lancet Infect Dis 2007;7:614–24.

[71] De Maria Jr A, Lett SM. Vaccine the village. Clin Infect Dis 2010;50:1346–8.[72] Skowronski DM, Janjua NZ, Tsafack EPS, Ouakki M, Hoang L, De Serres G. The

number needed to vaccinate to prevent infant pertussis hospitalization anddeath through parent cocoon immunization. Clin Infect Dis 2012;54:318–27.

[73] Halasa NB, O’Shea A, Shi JR, LaFleur BJ, Edwards KM. Poor immune responses toa birth dose of diphtheria, tetanus and acellular pertussis vaccine. J Pediatr2008;153:327–32.

[74] Knuf M, Schmitt HJ, Jacquet JM, Collard A, Kieninger D, Meyer CU, Siegrist CA,Zepp F. Booster vaccination after neonatal priming with acellular pertussisvaccine. J Pediatr 2010;156:675–8.

Paediatric Respiratory Reviews 16 (2015) 119–126

Review

Evidence compendium and advice on social distancing and otherrelated measures for response to an influenza pandemic

Harunor Rashid 1,*, Iman Ridda 1,2, Catherine King 1, Matthew Begun 3, Hatice Tekin 4,James G. Wood 3, Robert Booy 1,5

1 National Centre for Immunisation Research and Surveillance of Vaccine Preventable Diseases (NCIRS), The Children’s Hospital at Westmead, NSW 2145,

Australia2 School of Public Health, Tropical Medicine & Rehabilitation Sciences, James Cook University, Townsville, Australia3 School of Public Health and Community Medicine, Faculty of Medicine, The University of New South Wales, Sydney, NSW 2052, Australia4 School of Mathematics and Statistics, The University of Sydney, Australia5 Marie Bashir Institute for Infectious Diseases and Biosecurity, School of Biological Sciences and Sydney Medical School, The University of Sydney, Australia

A R T I C L E I N F O

Keywords:

Evidence compendium

Home isolation

Internal mobility restriction

Restriction of mass gatherings

Pandemic influenza

Quarantine

School Closure

Social distancing measures

Work-place related interventions.

S U M M A R Y

The role of social distancing measures in mitigating pandemic influenza is not precisely understood. To

this end, we have conducted a systematised review, particularly in light of the 2009 pandemic influenza,

to better inform the role of social distancing measures against pandemic influenza.

Articles were identified from relevant databases and the data were synthesised to provide evidence on

the role of school or work place-based interventions, case-based distancing (self-isolation, quarantine),

and restriction of mobility and mass gatherings.

School closure, whether proactive or reactive, appears to be moderately effective and acceptable in

reducing the transmission of influenza and in delaying the peak of an epidemic but is associated with very

high secondary costs. Voluntary home isolation and quarantine are also effective and acceptable

measures but there is an increased risk of intra-household transmission from index cases to contacts.

Work place-related interventions like work closure and home working are also modestly effective and are

acceptable, but likely to be economically disruptive. Internal mobility restriction is effective only if

prohibitively high (50% of travel) restrictions are applied and mass gatherings occurring within 10 days

before the epidemic peak are likely to increase the risk of transmission of influenza.

� 2014 Elsevier Ltd. All rights reserved.

EDUCATIONAL AIMS

The readers will become familiar with:

� Critical evaluation of the impact of social distancing measures against pandemic influenza in light of the latest evidence� Direct and indirect costs of implementing certain social distancing measures (e.g., school and workplace closures) in mitigating

pandemic influenza� Acceptability of social distancing measures among members of public and their expectations

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

* Corresponding author. National Centre for Immunisation Research and

Surveillance Kids Research Institute at The Children’s Hospital at Westmead Cnr

Hawkesbury Road and Hainsworth Street, Westmead Locked Bag 4001, Westmead,

NSW 2145, Australia. Tel.: +61 29845 1489; fax: +61 29845 1418.

E-mail address: harunor.rashid@health.nsw.gov.au (H. Rashid).

1526-0542/$ – see front matter � 2014 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.prrv.2014.01.003

INTRODUCTION

In an average season between 0.3 and 9.8% children presentwith influenza for medical attention; this incidence is greater inchildren under 5 years of age [1]. Globally, an estimated 90 million

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126120

cases of influenza occur each year in children aged under 5, ofwhich an estimated 20 million are associated with acute lowerrespiratory tract infection [2]. For pandemic strains the burden canbe greater, with elevated hospital and emergency departmentsadmission rates throughout the population [1], elevated deathrates and wider socioeconomic impacts [3]. Pandemic plans havebeen developed by organisations such as the WHO, individualcountries and local groups such as businesses, in an attempt tomitigate such risks. Amongst the public health measuresconsidered (e.g., vaccination, antiviral stock piling), those thatpromote reduced infectious contact in societies, or ‘‘socialdistancing’’ may complement other pandemic planning measuresin decreasing the likelihood and severity of pandemic influenzatransmission.

In contrast to technological interventions such as vaccinationand anti-viral drugs, published evidence of the efficacy oreffectiveness of individual measures for social distancing islimited. The European Centre for Disease Prevention and Control(ECDC) published a technical report on public health measures toreduce pandemic influenza impact [4] that identified gaps in ourunderstanding of the role of social distancing measures inpandemic planning. However, the 2009 H1N1 influenza pandemichas stimulated more research in this area, including clinical andepidemiological studies, and mathematical modeling. Pandemicplans in a number of countries, including Australia are currentlybeing revised in light of published reviews of their effectivenessand therefore, it is timely to identify and review the latest evidenceregarding the potential impacts of social distancing and relatedmeasures [5]. This review assessed the impact of the followingmeasures: proactive or reactive school closure, workplace closure,working from home, isolation of cases and/or contacts (quar-antine), internal mobility restriction and cancellation of massevents.

SEARCH STRATEGY

Database searching was undertaken by an experienced medicallibrarian (CK) in December 2012. A copy of the full search can beobtained by contacting the authors; in short, the followingdatabases/sources were searched - Ovid Medline (1946 toNovember Week 3 2012), Ovid Embase (1980 to December Week3 2012), Cochrane Library (Database of Systematic Reviews Issue12 of 12, December 2012, Database of Abstracts of Reviews ofEffects Issue 4 of 4, October 2012, Central Register of ControlledTrials Issue 12 of 12, December 2012, Health TechnologyAssessment Database Issue 4 of 4, October 2012, NHS EconomicEvaluation Database Issue 4 of 4, October 2012), SCOPUS (1823-December 2012), Web of Science (including Science Citation Index1900-December 2012) and consultation with national and inter-national guidelines.

Where available, database controlled vocabulary terms were‘exploded’ to include narrower terms and were used in combina-tion with relevant text-words. Terms representing all forms ofinfluenza (both seasonal and pandemic influenza) such as‘Influenza, Human’, ‘Influenza A Virus, H1N1 Subtype’ and‘Influenza A Virus, H5N1 Subtype’ were combined with termsrepresenting aspects of social distance and other mitigationmeasures. Searches were limited to ‘Human’ where this limitwas available. No language or abstract restrictions were applied.

A total of 3684 titles were retrieved; 134 studies were initiallyconsidered and 80 were finally included in this review. Emphasiswas given to studies published in, or after, 2008. This landmarkyear was chosen as: a) most of the major national or internationalguidelines (e.g., ECDC menu, Australian Health Management Planfor Pandemic Influenza [AHMPPI]) were published in or after 2008,and b) this allowed the evidence compendium to be updated in

light of the studies published on the 2009 pandemic. The ECDCtechnical report on pandemic influenza (ECDC menu) has beenused as a basic template for this review, allowing for quickcomparison to identify the differences and latest updates [4].

Studies involving mathematical modeling data were abstractedindividually by two authors (HR and MB) and subsequently cross-checked for consistency, and whenever appropriate, the extracteddata were merged, while the other studies were abstracted andchecked by a single author (HR or IR).

An arbitrary scale was used for effectiveness: ‘high’ to mean anoverall risk reduction of >50%, ‘moderate’ to mean a reductionbetween 10% and 50% and ‘mild’ to mean a reduction of <10%.Similarly, an arbitrary scale was also employed for economicimpact: ‘massive’ meant an impact of hundreds of millions orbillions of dollars, ‘major’ meant an impact in the range of millionsof dollars, ‘considerable’ meant an impact of hundred thousands ofdollars, and ‘moderate’ meant a smaller impact.

RESULTS

The effectiveness and limitations of each individual interven-tion are summarised in Tables 1 and 2, and described below.

SCHOOL CLOSURE

Evidence of effectiveness

Studies suggest that school closure, whether proactive orreactive, reduces transmission of influenza and delays theepidemic peak. The majority of modeling and observationalstudies suggest a reduction in influenza occurrence or transmis-sion following school closure but with wide variance (range 1-50%)[6–21]. Other studies, in which transmission between children isassumed to be very influential, have predicted effectiveness as highas 90-100% [22–24].

School closures may also delay the epidemic by a week or two,depending on the timing of the closure [19], while model-fits to2009 surveillance data from Alberta, Canada suggest that closuresdelayed the 2nd wave of the influenza pandemic [20].

Evidence from surveillance studies is limited: a 2007 study inWashington State, USA used school absenteeism as a proxy forseasonal influenza activity in examining the effect of a week-longwinter school closure but found no difference between schools onbreak and those remaining in session in post-break absenteeism[10]. A Bangkok study from mid-2009 concluded that the incidenceof pandemic influenza declined following school closure in affectedchildren and adolescents aged 6-20 years but increased amongyoung children (aged <6 years) and in young adults (20-40 years).However, the study did not report statistical support of theseconclusions [25].

Reduced transmission is assumed to stem from a decreasednumber of contacts by schoolchildren during school closure; forinstance, surveys conducted in the UK reported a 50-65% reductionin students’ typical daily number of contacts [13,26,27]. Reduc-tions were greatest for contacts within the school, e.g., approxi-mately 80% reduction in contacts with classmates and ‘year mates’[27].

However, the optimum timing of school closure depends on theduration of closure. For instance, a simulation study showed thatfor short duration (< 6 weeks) school closures it is more effective tobegin closure once infection is already widespread; while forlonger durations (> 6 weeks), beginning as early as reasonable issuperior [28].

Some modeling studies have found smaller effects from schoolclosure due to shifting of student’s contacts to other groups duringthe closure period; of note, in surveys conducted in the USA, UK

Table 2Summary of isolation, quarantine, travel restriction and cancellation of mass events

Interventions Effectiveness Disadvantages Practicalities

Self-isolation of cases There are limited data, overall

effectiveness of the measure is

moderate; may delay the peak of

influenza when combined with

other measures.

Those who share a room (household members,

roommates) with the index case may be at risk

of acquiring the infection.

Although likely to be acceptable, compliance

with self-isolation may not be uniform across

the population. The period a person needs to be

isolated is not precisely defined and there are

variations.

Quarantine of contacts Modelling studies show that

quarantine decreases peak case

load, attack rate, and delays the

peak.

Home isolation of contacts sharing the same

room or facilities with index cases significantly

increases the risk of acquiring the infection

among the contacts.

Those affected by isolation and quarantine are

likely to report distress due to fear and risk

perceptions.

Mobility restriction Modelling studies suggest that a

high travel restriction (50%) delays

the peak of influenza. A minimal

travel restriction is not helpful.

Restrictions on travelling will halt trade and

business and, may indirectly impair supply of

essential commodities and disrupt economic

activities.

In places where travel is frequent imposing

travel restrictions would be impractical.

Cancellation of mass events Effectiveness is not proven but may

be of theoretical benefit if cancelled

around the peak of the epidemic.

Secondary effects are major especially on those

who organise meetings and events and derive

an income from that work. Cancellation of

certain mass gatherings (e.g., religious events)

could raise profound questions regarding both

faith and family.

Cancellation of mass events is not always

straightforward and may be associated with

practical problems, but major events can be

organised in the midst of a pandemic by taking

stringent containment measures.

Table 1Summary of school-, and work-based interventions

Interventions Effectiveness Disadvantages Practicalities

Proactive school closure Reduction in influenza

transmission from 1% to 50%.

Delays the peak of the epidemic

by a week or two.

The secondary economic and social effects of

proactive school closures are massive. Based on

the UK data economic modeling shows that

0.2–1% of GDP loss occurs due to school closure

lasting the duration of a pandemic wave.

The purpose of implementing this policy may

not be achieved as a large proportion of

children (up to 70%) may engage in outdoor

activities during the closure period. Also, there

may be considerable variation in decisions

about school closure.

Reactive school closure Reactive school closures may

reduce the transmission of

influenza by about 7-15%,

rarely up to 90-100%.

As a large proportion (16% to 45%) of parents

would remain absent from work to take care of

children there will be massive economic effect

resulting from income loss. A US study suggests

that closing schools for 8 weeks would result in

median net costs of $21.0 billion just for the

state of Pennsylvania. In places where children

are offered free/reduced cost school meals they

will miss out on the meals.

Modelling studies show that any type of school

closure may need to be maintained throughout

most of the epidemic (i.e., at least 8 weeks) to

have any significant effect. Practicalities of

keeping schools closed for a long duration are

considerable as the secondary effects will be

magnified. Up to 69% children may gather for

social activities at least once during the closure

so the purpose of closing will not be optimally

achieved.

Workplace closure Modeling study suggests that

10% work place closure has only

modest impact while 33%

workplace closure lessens the

attack rate to less than 5%, and

delays the peak by 1 week.

Workplace closures economically affect

workplaces through decreased business

productivity and employees through loss of

workdays and income.

Although workplace closure is a highly popular

intervention, the practicalities are likely to be

considerable, as some businesses and

organisations could not close completely and

for an extended period. Also, a fairly large and

unwieldy proportion (at least one third) of

workplace closures would be necessary to bring

an influenza epidemic under control.

Home working It is moderately effective in

reducing transmission of

influenza by about 20% to 30%.

May not be equally feasible for all, especially for

self-employed people who may suffer serious

financial problems.

With the use of latest technology home

working is possible for service sectors but not

for factories with physical outputs.

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126 121

and Australia, 40- 70% of students took part in outdoor activities atleast once during school closures [27,29–31].

Secondary effects

The secondary economic and social effects of school closures arepotentially massive [4]. Studies from the USA, Australia andTaiwan suggest that 16-45% of parents would need to take leave tosupervise children at home [29–35], 16-18% of parents would loseincome, [31,34] and about 20% of households would have difficultyarranging childcare [29,33,36].

In the UK, health economists have estimated that the cost ofschool closure can be significant, at £0.2 billion – £1.2 billion perweek, amounting to around 0.2–1% of GDP for school closure overa pandemic wave (around 12 weeks) [37], with comparableresults in terms of cost found in an Australian modeling study [7].

An analysis for the state of Pennsylvania, USA suggested themedian net cost of closing schools for 8 weeks is estimated to be$21.0 billion (>3% GDP) although details of the loss of earningscalculation were not transparent [38]. A comparative analysis ofpublic health measures for pandemic mitigation found thatcontinuous universal school closure was less cost-effective thanstockpiling antiviral drugs or pre-pandemic vaccines, with schoolclosure 14 or 21 times more costly for equivalent morbidity andmortality benefits than intervention strategies with antivirals orpre-vaccination respectively. However, this corresponds toclosing of all schools for 26 weeks [39], with other closurestrategies likely to be more efficient. School closures may alsohave additional social effects as there is a risk of children being leftwithout care or in the care of under-aged siblings [40] andinability to access aid services such as free/reduced cost schoolmeals [33,36].

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126122

Practicalities and expectations

There may be considerable variation in decisions about schoolclosure often leading to disagreements between school and publichealth officials on issues arising from laws mandating the number ofinstruction days schools must provide to receive state funding, orschool authorities’ fear of risking this funding or incurring extra costsby extending the school year [41]. Such variation also exists in legalframeworks relating to school closure [42,43]. Significant transmis-sion among school pupils (and others) may also occur outside of classthrough group activities such as choirs or through parties and otherrecreational events [44]. This inability to interrupt all forms oftransmission between students suggests that early concurrent use ofmultiple public health measures might be more desirable [45].Modeling studies show that maintenance of school closure through-out most of the epidemic (i.e., at least 8 weeks) may be required tosignificantly affect the overall attack rate, but that short-term closureis likely to be more cost-effective [7,46,47].

There are also questions as to public compliance with schoolclosure. While surveys have suggested that most families (73% to97%) would accept the dismissal decision, only half of surveyedfamilies would keep children at home [29,33,34,48,49]. Othersurveys disclosed that about 61% of households would find schoolclosure to be difficult, [36] 3% would experience a major problemwhile only 20% thought dismissal was a minor problem [33].

CHANGES TO WORK ARRANGEMENTS

Evidence of effectiveness

We found only 3 empirical studies examining the impact ofchanges to work arrangements on influenza transmission risks, witha US study during 2009-10 finding reduced risks of self-reported ILIamong individuals who could work from home (8.2% versus 12.8%,p<0.05) or stay home from work for 7-10 days (8.8% vs 12.2%p<0.05) compared with those who could not [50]. In a quasi-clusterrandomised trial conducted in Japan in 2009, the workplace policy ofbeing able to remain at home on full pay produced a statisticallysignificant reduction in employee risk of H1N1 pandemic influenza(RR 0.8 (0.66-0.97), p=0.023) [51]. However, workers required tostay home with sick household members were at increased risk ofH1N1 infection (RR 2.17 (1.48-3.18), p<0.001). Finally, a prospectiveUS study found that prior teleworking arrangements reduced by 30%the rate of employees attending work with severe ILI symptoms(p=0.026), with potential for reduced transmission [52].

Analysis through modeling studies was also limited, with asingle US study suggesting that attack rates can be much reducedby workplace closures, with epidemic attack rates declining from18.6% to 11.9% with 10% workplace closure (WC), and from 18.6% to4.9% with 33% WC; a shift in peak by approximately 1 weekoccurred in the latter case [53]. However, this study assumes 100%school closure in addition to workplace closures and representsimpacts of a combined intervention.

Secondary effects

Workplace closures will affect employees through loss ofworkdays and income, and consumers through loss of essential andimportant goods [1,7]. A US survey identified self-employedindividuals and those unable to work from home as at heightenedrisk of serious financial difficulties if isolated from work [54].

Practicalities and expectations

Practical obstacles to workplace closure are considerable,including the inability of some workplaces to close for an extended

period, the large scale of closures needed for epidemic control [53]and the need for some businesses to increase operations in apandemic (e.g., medical supplies). Technological solutions (e.g.,broadband internet access, virtual private networks, videoconfer-encing) allow execution of many key functions from home, with a2006 survey of the Australian public health workforce indicatingthat 73% of staff would be able to work from home in the event ofschool closure requiring them to supervise their children at home[32].

Home working (especially if on full pay) or staying home fromwork is likely to be highly acceptable [51,54], but prolongedclosures could lead to concerns about job security and placeeconomic strains on families [55].

CASE-BASED DISTANCING

Evidence of effectiveness

There were limited studies evaluating these measures, withmoderate effectiveness found in modeling simulations. A Japanesemodeling study suggested that total community infections wouldfall as the proportion of children or adults isolated at homeincreased [16], with support from other modeling studies [56,57].

Simulations of quarantining household contacts of index casesalso suggests benefit [19,58,59]. In a modeling study applied toMongolia (a useful example in terms of its isolation), quarantining50% of all case contacts over a period of 4 weeks prior to theepidemic peak, reduced the peak case-load and attack rate by 25%,and 1.5% respectively and delayed the peak by around 1 week [19].Another modeling study calibrated to Asian influenza data of 1957-58, predicts reductions in rates of illness and mortality by about50% if cases of ILI and their household contacts stayed home [58].

Secondary effects

Secondary effects of self-isolation are moderate, with house-hold contacts of the index case at risk of acquiring the infection. Across-sectional study conducted in Australia identified secondarytransmission to 18 of 122 (15%) susceptible household contacts ofindex cases [60], while an observational study in Chineseuniversity students found a significant increase in rates of ILI forstudents quarantined in a room with a confirmed case of H1N1pandemic influenza over those not exposed to a H1N1 case (26%versus 6%, p=0.02) [61].

People affected by isolation and quarantine are likely to reportdistress due to fear and risk perceptions, especially if clearguidelines on minimising risks of infection are lacking [62]. In a USsurvey, 28% of employed respondents reported that they would belikely to lose their job or business if required to stay home fromwork for 7–10 days in the event of a pandemic influenza outbreak,with this 5 times more likely among respondents who would notreceive paid leave [54].

Practicalities and expectations

There are a number of practical difficulties associated withimposing quarantine and isolation. Firstly, a viral kinetics studyhas shown that about a quarter of patients shed virus beyond thestandard isolation period of seven days (or until resolution of feverif longer) [63,64], suggesting a possible need for review of thisperiod. Secondly, consideration needs to be given to supporting thepsychological, financial, social, physical and other needs of patients[65], particularly with respect to the adverse impacts on contactsexposed through quarantine to index cases [61]. Justification forcoercive social distancing measures, like isolation and quarantine,involves competing ethical priorities which must be balanced with

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126 123

consideration of triggers for such measures to be enacted, andevidence of effect [66].

Underpinning effective public health implementation of isola-tion and quarantine during pandemic influenza are tools such as‘flexible incident control systems’ and multi-user access electronicdatabases [67], which may not be available in resource poorsettings. Even with such tools, quarantine and isolation areresource intensive processes that may only be feasible when casenumbers are low. In addition, effectiveness depends on a highproportion of cases being identified, which was shown not to bethe case for H1N1 in 2009.

Acceptability of self-isolation and quarantine is variable butgenerally high, with surveys conducted in the USA and Australiaduring the 2009 pandemic, finding >80% people willing to stayhome from work or school [68,69], while 53-76% people werewilling to perform self-isolation [68,70]. Regarding quarantine, inAustralian surveys about 85% to 90% of households reportedadhering to, or willingness to comply with the requirement to stayat home (98% in those respondents who were provided with anexplanation about pandemic influenza) [48,71]; while a US studyfound strong public support for use of quarantine and serious legalsanctions against non-compliers [72]. However, in a survey amongstudents and faculty members of a US university during the 2009pandemic, only 6.4% students and 8.6% (5 of 58) faculty with acuterespiratory infection reported staying home while ill [73],suggesting compliance may not be uniform across the population.However, concerned students and faculties were more adherentthan their unconcerned counterparts, with similar findings in anAustralian survey comparing well-informed households to ill-informed households [70].

RESTRICTIONS ON MOVEMENT AND GATHERINGS

Evidence of effectiveness

Modeling studies comprise most recent research in this area,although a systematic review suggests some mass gatheringsare associated with increased influenza transmission, whilearchival studies of the 1918-19 pandemic imply that restrictingmass gatherings was beneficial, particularly with early imple-mentation [74]. A modeling paper suggested that mass gather-ings shortly before the epidemic peak could increase the peakheight by about 10% but at other times, the impact would besmall.[75]

Simulations suggest that moderate delays (1-1.5 weeks) couldbe achieved by strict internal mobility restrictions (>50% reductionin mean travel frequency) applied in the early stages of thepandemic for a period of 2-4 weeks [19]. The peak ILI could also fallby about 12% with these restrictions, with similar results in anadditional modeling study [53]. Another modeling study suggeststhat weak travel restrictions (10% travel restrictions) mightactually increase attack rates due to prevented travelers transmit-ting more infections within their local area [53]. A South Koreansimulation study suggests that travel restrictions delay the spreadof the disease into new cities, but increase the peak height ofinfected population in all cities [76]. Thus, there is not a universalconsensus from modeling studies on the impact of restrictions onmovement.

Secondary effects

We did not find new evidence of these impacts but the SARSexperience provided an example of potential impacts, which arealso summarised in the ECDC report, and appear likely to have amajor economic impact, particularly on sectors such as tourismand hospitality [4].

Practicalities and expectations

It may be impractical to impose travel restrictions in countrieswith a high level of internal travel. Historical data from the 1918-19 pandemic suggest that in places where travel was frequent,restrictions were not effective [77]. Acceptability may varybetween settings, with surveys conducted in the USA, Argentina,Mexico, UK, Japan reporting that 11-54% of respondents avoidedtravelling long distances by aeroplane, train, or bus during the2009 pandemic [78], while 97% of Australians reported willingnessto avoid air travel for one month if requested during a potentialpandemic [48].

Cancellation of meetings and events is not without precedentand contingency plans are generally in place [4]. However,cancellation may not be necessary if rigorous containmentmeasures are applied at mass gatherings, with Lim et al. high-lighting the inaugural Asian Youth Games in Singapore in 2009 as apositive example of control during an influenza pandemic [79].Acceptability for cancelling mass events may depend on the typeand significance of the gathering. In an Australian phone surveymore than 90% of individuals reported willingness to avoid publicevents during a pandemic influenza [48], while a US qualitativestudy revealed potential opposition to closure of religiousgatherings [55].

DISCUSSION

Overall effectiveness and quality of the evidence

A large number of published studies examining social distan-cing have been conducted since the 2009 pandemic of influenza.Overall, social distancing measures are found to be moderatelyeffective.

School or workplace closures and case-based measures such asisolation and quarantine of contacts are found to be at leastmodestly effective, but are likely to be costly and disruptive.Mobility restrictions may not be effective unless prohibitively highrestrictions are applied and are likely to be impractical fordeveloped countries that depend heavily on internal transport. Theeffect of other measures (e.g., cancellation of mass gatherings) isonly poorly established and requires further assessment.

Overall, the quality of the evidence was quite weak, drawingprimarily on observational or simulated data. In terms of moreestablished methods, we found one systematic review thatexamined the impact of restricting mass gatherings on influenzatransmission and one quasi-randomised trial assessing the effect ofhomeworking (on full pay) on influenza transmission which foundthe intervention to be moderately beneficial and highly acceptable[51].

School closure

Proactive or reactive school closure was considered to bemoderately effective but with high variance between studies. Thismeasure can reduce total infections, and delay and flatten theepidemic peak [9,21] but is likely to have massive economicimpacts with potential for much social disruption. Detailedeconomic analysis performed in the UK, and elsewhere, foundthat potential costs from school closure far outweigh savings fromprevented cases [37–39]. Several studies suggest that withoutmeasures to reduce out-of-school contacts, school closures wouldnot have a considerable effect on pandemic influenza [10,58], withan Australian survey noting that 74% of students participated inoutdoor activities more than once during school closure [29].

A US-based modeling study suggested that school closure mayneed to be maintained for at least 8 weeks to have significant

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126124

effects on overall attack rates, and that short closures of 2 weeks orless could be counterproductive and increase overall attack ratesby returning susceptible students to schools as the epidemic peaks[46]. However, other studies had conflicting results with aneconomic analysis of school closure as the sole intervention,suggesting that limited duration closure was more cost-effectivethan continuous closure [7]. In view of these limitations wesuggest that school closure be considered only for severepandemics.

Additional points for consideration in relation to school closureare: a) school closure alone may not be effective unless cessation ofother activities or gatherings (e.g., choirs, parties) is also ensured;[44] b) entire school system closures are unlikely to be moreeffective than individual school closures, so coordinated schoolclosure in jurisdictions should be discouraged unless hospitaldemand exceeds capacity [7,46,80]; c) defined triggers forimposing proactive or reactive school closure may have value[43,81]; d) full closure of school (not just dismissal of a class/classes) should be practiced [47]; and e) a transparent andconsistent strategy should be followed in communicating deci-sions on school closure to school authorities, the public and otherstakeholders [42].

Workplace-related interventions

The effectiveness of workplace closure or home workingstrategies were examined in several studies [50–53], with thesesuggesting modest effectiveness and high acceptability[48,51,52,68], especially if full payment is provided to employeesduring their absence from work [51]. However, high levels ofworkplace closure (at least 33%) combined with 100% schoolclosure was required in a modeling study to mitigate a pandemic[53].

Case-based distancing

Isolation of infected cases and quarantine of contacts may beeffective measures, but may be impractical once epidemics arewell-established [16,19,56,58]. Public acceptance of voluntaryisolation and quarantine in some countries is high [48,69,71] andin many countries legal mechanisms to restrict movement ofindividuals with ‘quarantinable’ diseases exist [65], but forcedisolation is unlikely to be practical or effective in most modernsocieties.

In relation to these measures, several points need to beconsidered: firstly, infection risks in suspected contacts could beelevated if ‘roomed in’ or sharing facilities (e.g., toilet, shower) withindex cases [61]; secondly, physical, psychological and social needsof the isolated or quarantined individuals should be addressed[82]; thirdly, the current standard duration of isolation (7 days oruntil becomes symptom free) may need to be revised toaccommodate patients that shed the virus for longer periods [63].

Restrictions on movement and gatherings

Interventions to restrict internal travel may not significantlyreduce influenza transmission unless a high proportion (at least50%) of restriction is applied [53]. Acceptability is likely to vary bycountry and mode of transport, and practicality is questionable inmost industrialised countries [48].

Some evidence suggests that mass gatherings can amplifyinfluenza transmission, especially if the event takes place shortlybefore the epidemic peak but available data provide little supportfor reduced transmission from cancelling mass gatherings [74,75].If requested, survey participants expressed willingness to avoidmass events particularly if the effects of pandemic influenza were

explained [48]. Gatherings for spiritual or other reasons mayprovide comfort in times of crisis and examples of safely organisingmass gatherings through rigorous control measures during andinfluenza pandemic have been described [78].

CONCLUSION

Since the 2009 pandemic, more published evidence on theeffectiveness of social distancing measures in mitigating pandemicinfluenza has appeared. Overall, social distancing measures appearmodestly effective and many are likely to be acceptable in the shortterm, but there is a lack of strong evidence.

FUTURE RESEARCH DIRECTIONS

� As there is a dearth of high quality studies, controlled trialsshould be prioritised in this field� The evidence compendium needs to be updated periodically to

incorporate emerging evidence� Systematic reviews on the interaction between individual

measures should be undertaken, including wider interventionslike face masks, vaccines and antivirals

Acknowledgement

This study was funded by the Australian Department of Healthobtained through a tender process (RFQ No: DoHA/130/1213).Iman Ridda holds an NHMRC Early Career Fellowship (630739) andJames Wood has received partial salary support from NHMRC CRE1021963.

References

[1] Ruf BR, Knuf M. The burden of seasonal and pandemic influenza in infants andchildren. Eur J Pediatr 2013 (Epub ahead of print)

[2] Nair H, Brooks WA, Katz M, et al. Global burden of respiratory infections due toseasonal influenza in young children: a systematic review and meta-analysis.Lancet 2011;378:1917–30.

[3] Marchisio P, Baggi E, Bianchini S, Principi N, Esposito S. Clinical and socio-economic impact of pediatric seasonal and pandemic influenza. Hum VaccinImmunother 2012;8:17–20.

[4] Guide to public health measures to reduce the impact of influenza pandemicsin Europe: ‘The ECDC Menu’; Stockholm, September 2009.(Available from:http://www.ecdc.europa.eu/en/publications/Publications/0906_TER_Publi-c_Health_Measures_for_Influenza_Pandemics.pdf).

[5] Kelly HA, Priest PC, Mercer GN, Dowse GK. We should not be complacent aboutour population-based public health response to the first influenza pandemic ofthe 21st century. BMC Public Health 2011;11:78.

[6] Orellano PW, Grassi A, Reynoso JI, Palmieri A, Uez O, Carlino O. [Impact ofschool closings on the influenza A (H1N1) outbreak in Tierra del Fuego,Argentina]. Rev Panam Salud Publica 2010;27:226–9.

[7] Halder N, Kelso JK, Milne GJ. Cost-effective strategies for mitigating a futureinfluenza pandemic with H1N1 2009 characteristics. PLoS ONE 2011;6:e22087.

[8] Egger JR, Konty KJ, Wilson E, et al. The effect of school dismissal on rates ofinfluenza-like illness in New York City schools during the spring 2009 novelH1N1 outbreak. J Sch Health 2012;82:123–30.

[9] Cauchemez S, Ferguson NM, Wachtel C, et al. Closure of schools during aninfluenza pandemic. Lancet Infect Dis 2009;9:473–81.

[10] Rodriguez CV, Rietberg K, Baer A, Kwan-Gett T, Duchin J. Association betweenschool closure and subsequent absenteeism during a seasonal influenzaepidemic. Epidemiology 2009;20:787–92.

[11] Heymann AD, Hoch I, Valinsky L, Kokia E, Steinberg DM. School closure may beeffective in reducing transmission of respiratory viruses in the community.Epidemiol Infect 2009;137:1369–76.

[12] Wheeler CC, Erhart LM, Jehn ML. Effect of school closure on the incidence ofinfluenza among school-age children in Arizona. Public Health Rep2010;125:851–9.

[13] Eames KTD, Tilston NL, White PJ, Adams E, Edmunds WJ. The impact of illnessand the impact of school closure on social contact patterns. Health TechnolAssess 2010;14:267–312.

[14] Chowell G, Echevarria-Zuno S, Viboud C, et al. Characterizing the epidemiologyof the 2009 influenza A/H1N1 pandemic in Mexico. PLoS Med2011;8:e1000436.

[15] Yu H, Cauchemez S, Donnelly CA, et al. Transmission dynamics, border entryscreening, and school holidays during the 2009 influenza A (H1N1) pandemic,China. Emerg Infect Dis 2012;18:758–66.

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126 125

[16] Yasuda H, Suzuki K. Measures against transmission of pandemic H1N1 influ-enza in Japan in 2009: simulation model. Euro Surveill 2009;14. pii=19385.

[17] Wu JT, Cowling BJ, Lau EH, et al. School closure and mitigation of pandemic(H1N1) 2009, Hong Kong. Emerg Infect Dis 2010;16:538–41.

[18] Hens N, Ayele GM, Goeyvaerts N, et al. Estimating the impact of school closureon social mixing behaviour and the transmission of close contact infections ineight European countries. BMC Infect Dis 2009;9:187.

[19] Bolton KJ, McCaw JM, Moss R, et al. Likely effectiveness of pharmaceutical andnon-pharmaceutical interventions for mitigating influenza virus transmissionin Mongolia. Bull World Health Organ 2012;90:264–71.

[20] Earn DJ, He D, Loeb MB, Fonseca K, Lee BE, Dushoff J. Effects of school closure onincidence of pandemic influenza in Alberta, Canada. Ann Intern Med2012;156:173–81.

[21] Cauchemez S, Valleron AJ, Boelle PY, Flahault A, Ferguson NM. Estimating theimpact of school closure on influenza transmission from Sentinel data. Nature2008;452:750–4.

[22] Sypsa V, Hatzakis A. School closure is currently the main strategy to mitigateinfluenza A(H1N1)v: a modeling study. Euro Surveill 2009;14. pii= 19240.

[23] Glass RJ, Glass LM, Beyeler WE, Min HJ. Targeted social distancing design forpandemic influenza. Emerg Infect Dis 2006;12:1671–81.

[24] Kawaguchi R, Miyazono M, Noda T, Takayama Y, Sasai Y, Iso H. Influenza(H1N1) 2009 outbreak and school closure, Osaka Prefecture. Japan Emerg InfectDis 2009;15:1685.

[25] Chieochansin T, Makkoch J, Suwannakarn K, Payungporn S, Poovorawan Y.Novel H1N1 2009 influenza virus infection in Bangkok, Thailand: effects ofschool closures. Asian Biomed 2009;3:469–75.

[26] Eames KT, Tilston NL, Edmunds WJ. The impact of school holidays on the socialmixing patterns of school children. Epidemics 2011;3:103–8.

[27] Jackson C, Mangtani P, Vynnycky E, et al. School closures and student contactpatterns. Emerg Infect Dis 2011;17:245–7.

[28] Zhang T, Fu X, Kwoh CK, et al. Temporal factors in school closure policy formitigating the spread of influenza. J Public Health Policy 2011;32:180–97.

[29] Effler PV, Carcione D, Giele C, Dowse GK, Goggin L, Mak DB. Householdresponses to pandemic (H1N1) 2009-related school closures, Perth, WesternAustralia. Emerg Infect Dis 2010;16:205–11.

[30] Gift TL, Palekar RS, Sodha SV, et al. Household effects of school closure duringpandemic (H1N1) 2009, Pennsylvania, USA. Emerg Infect Dis 2010;16:1315–7.

[31] Centers for Disease C. Prevention. Impact of seasonal influenza-related schoolclosures on families - Southeastern Kentucky, February 2008. MMWR MorbMortal Wkly Rep 2009;58:1405–9.

[32] Dalton CB, Durrheim DN, Conroy MA. Likely impact of school and childcareclosures on public health workforce during an influenza pandemic: a survey.Commun Dis Intell Q Rep 2008;32:261–2.

[33] Centers for Disease C. Prevention. Parental attitudes and experiences duringschool dismissals related to 2009 influenza A (H1N1) — United States, 2009.MMWR Morb Mortal Wkly Rep 2010;59:1131–4.

[34] Chen WC, Huang AS, Chuang JH, Chiu CC, Kuo HS. Social and economic impactof school closure resulting from pandemic influenza A/H1N1. J Infect2011;62:200–3.

[35] Borse RH, Behravesh CB, Dumanovsky T, et al. Closing schools in response tothe 2009 pandemic influenza A H1N1 virus in New York City: economic impacton households. Clin Infect Dis 2011;52(Suppl 1):S168–72.

[36] Jarquin VG, Callahan DB, Cohen NJ, et al. Effect of school closure from pan-demic (H1N1) 2009, Chicago, Illinois, USA. Emerg Infect Dis 2011;17:751–3.

[37] Sadique MZ, Adams EJ, Edmunds WJ. Estimating the costs of school closure formitigating an influenza pandemic. BMC Public Health 2008;8:135.

[38] Brown ST, Tai JH, Bailey RR, et al. Would school closure for the 2009 H1N1influenza epidemic have been worth the cost?.: a computational simulation ofPennsylvania. BMC Public Health 2011;11:353.

[39] Sander B, Nizam A, Garrison Jr LP, Postma MJ, Halloran ME, Longini Jr IM.Economic evaluation of influenza pandemic mitigation strategies in the UnitedStates using a stochastic microsimulation transmission model. Value Health2009;12:226–33.

[40] Berkman BE. Mitigating pandemic influenza: the ethics of implementing aschool closure policy. J Public Health Manag Prac 2008;14:372–8.

[41] Klaiman T, Kraemer JD, Stoto MA. Variability in school closure decisions inresponse to 2009 H1N1: a qualitative systems improvement analysis. BMCPublic Health 2011;11:73.

[42] Stern AM, Cetron MS, Markel H. Closing the schools: lessons from the 1918-19U.S. influenza pandemic. Health Aff (Millwood) 2009;28:w1066–78.

[43] Zhang T, Fu X, Ma S, et al. Evaluating temporal factors in combined interven-tions of workforce shift and school closure for mitigating the spread ofinfluenza. PLoS One 2012;7:e32203.

[44] Kar-Purkayastha I, Ingram C, Maguire H, Roche A. The importance of schooland social activities in the transmission of influenza A(H1N1)v: England, April- June 2009. Euro Surveill 2009;14. pii=19311.

[45] Koonin LM, Cetron MS. School closure to reduce influenza transmission. EmergInfect Dis 2009;15:137–8.

[46] Lee BY, Brown ST, Cooley P, et al. Simulating school closure strategies tomitigate an influenza epidemic. J Public Health Manag Pract 2010;16:252–61.

[47] Uchida M, Tsukahara T, Kaneko M, Washizuka S, Kawa S. Effect of short-termschool closures on the H1N1 pandemic in Japan: a comparative case study.Infection 2012;40:549–56.

[48] Eastwood K, Durrheim D, Francis JL, et al. Knowledge about pandemic influ-enza and compliance with containment measures among Australians. BullWorld Health Organ 2009;87:588–94.

[49] Horney JA, Moore Z, Davis M, MacDonald PD. Intent to receive pandemicinfluenza A (H1N1) vaccine, compliance with social distancing and sources ofinformation in NC, 2009. PLoS ONE 2010;5:e11226.

[50] Kumar S, Quinn SC, Kim KH, Daniel LH, Freimuth VS. The impact of workplacepolicies and other social factors on self-reported influenza-like illness inci-dence during the 2009 H1N1 pandemic. Am J Public Health 2012;102:134–40.

[51] Miyaki K, Sakurazawa H, Mikurube H, et al. An effective quarantine measurereduced the total incidence of influenza A H1N1 in the workplace: anotherway to control the H1N1 flu pandemic. J Occup Health 2011;53:287–92.

[52] Rousculp MD, Johnston SS, Palmer LA, Chu BC, Mahadevia PJ, Nichol KL.Attending work while sick: Implication of flexible sick leave policies. J OccupEnviron Med 2010;52:1009–13.

[53] Mao L. Evaluating the combined effectiveness of influenza control strategiesand human preventive behavior. PLoS ONE 2011;6:e24706.

[54] Blake KD, Blendon RJ, Viswanath K. Employment and compliance with pan-demic influenza mitigation recommendations. Emerg Infect Dis 2010;16:212–8.

[55] Baum NM, Jacobson PD, Goold SD. Listen to the people’’: public deliberationabout social distancing measures in a pandemic. Am J Bioeth 2009;9:4–14.

[56] Lee S, Chowell G, Castillo-Chavez C. Optimal control for pandemic influenza:the role of limited antiviral treatment and isolation. J Theor Biol2010;265:136–50.

[57] Velasco-Hernandez JX, Leite MC. A model for the A(H1N1) epidemic in Mexico,including social isolation. Salud Publica Mex 2011;53:40–7.

[58] Haber MJ, Shay DK, Davis XM, et al. Effectiveness of interventions to reducecontact rates during a simulated influenza pandemic. Emerg Infect Dis2007;13:581–9.

[59] Navarro-Robles E, Martinez-Matsushita L, Lopez-Molina R, Fritz-Hernandez J,Flores-Aldana BA, Mendoza-Perez JC. [Model to estimate epidemic patterns ofinfluenza A (H1N1) in Mexico]. Rev Panam Salud Publica 2012;31:269–74.

[60] van Gemert C, Hellard M, McBryde ES, et al. Intrahousehold transmission ofpandemic (H1N1) 2009 virus, Victoria, Australia. Emerg Infect Dis2011;17:1599–607.

[61] Chu CY, Li CY, Zhang H, et al. Quarantine methods and prevention of secondaryoutbreak of pandemic (H1N1) 2009. Emerg Infect Dis 2010;16:1300–2.

[62] Johal SS. Psychosocial impacts of quarantine during disease outbreaks andinterventions that may help to relieve strain. N Z Med J 2009;122:47–52.

[63] Chin BS, Chae YT, Choi HK, et al. Viral shedding of 2009 pandemic H1N1 andevaluation of quarantine recommendations. Jpn J Infect Dis 2012;65:105–10.

[64] Australian Government Department of Health and Ageing. Australian HealthManagement Plan for Pandemic influenza, 2008.(Available from: http://www.flupandemic.gov.au/internet/panflu/publishing.nsf/content/8435EDE93CB6FCB8CA2573D700128ACA/$File/Pandemic%20FINAL%20web-ready.pdf).

[65] Bennett B. Legal rights during pandemics: federalism, rights and public healthlaws–a view from Australia. Public Health 2009;123:232–6.

[66] Selgelid MJ. Pandethics. Public Health 2009;123:255–9.[67] Binns PL, Sheppeard V, Staff MP. Isolation and quarantine during pandemic

(H1N1) 2009 influenza in NSW: the operational experience of public healthunits. N S W Public Health Bull 2010;21:10–5.

[68] Loustalot F, Silk BJ, Gaither A, et al. Household transmission of 2009 pandemicinfluenza A (H1N1) and nonpharmaceutical interventions among householdsof high school students in San Antonio, Texas. Clin Infect Dis 2011;52(Suppl1):S146–53.

[69] Teh B, Olsen K, Black J, et al. Impact of swine influenza and quarantinemeasures on patients and households during the H1N1/09 pandemic. ScandJ Infect Dis 2012;44:289–96.

[70] Kavanagh AM, Bentley RJ, Mason KE, et al. Sources, perceived usefulness andunderstanding of information disseminated to families who entered homequarantine during the H1N1 pandemic in Victoria, Australia: a cross-sectionalstudy. BMC Infect Dis 2011;11:2.

[71] McVernon J, Mason K, Petrony S, et al. Recommendations for and compliancewith social restrictions during implementation of school closures in the earlyphase of the influenza A (H1N1) 2009 outbreak in Melbourne, Australia. BMCInfect Dis 2011;11:257.

[72] Tracy CS, Rea E, Upshur RE. Public perceptions of quarantine: community-based telephone survey following an infectious disease outbreak. BMC PublicHealth 2009;9:470.

[73] Mitchell T, Dee DL, Phares CR, et al. Non-pharmaceutical interventions duringan outbreak of 2009 pandemic influenza A (H1N1) virus infection at a largepublic university, April-May 2009. Clin Infect Dis 2011;52(Suppl 1):S138–45.

[74] Ishola DAJPN. Could influenza transmission be reduced by restricting massgatherings? Towards an evidence-based policy framework. Epidemiol GlobHealth 2011;1:33–60.

[75] Shi P, Keskinocak P, Swann JL, Lee BY. The impact of mass gatherings andholiday traveling on the course of an influenza pandemic: a computationalmodel. BMC Public Health 2010;10:778.

[76] Lee JM, Choi D, Cho G, Kim Y. The effect of public health interventions on thespread of influenza among cities. J Theor Biol 2012;293:131–42.

[77] Markel H, Stern AM, Navarro JA, Michalsen JR, Monto AS, DiGiovanni C.Nonpharmaceutical influenza mitigation strategies, US communities, 1918-1920 pandemic. Emerg Infect Dis 2006;12:1961–4.

[78] SteelFisher GK, Blendon RJ, Ward JRM, Rapoport R, Kahn EB, Kohl KS. Publicresponse to the 2009 influenza A H1N1 pandemic: A polling study in fivecountries. Lancet Infect Dis 2012;12:845–50.

H. Rashid et al. / Paediatric Respiratory Reviews 16 (2015) 119–126126

[79] Lim HC, Cutter J, Lim WK, Ee A, Wong YC, Tay BK. The influenza A (H1N1-2009)experience at the inaugural Asian Youth Games Singapore 2009: mass gather-ing during a developing pandemic. Br J Sports Med 2010;44:528–32.

[80] House T, Baguelin M, Van Hoek AJ, et al. Modelling the impact of local reactiveschool closures on critical care provision during an influenza pandemic. ProcBiol Sci 2011;278:2753–60.

[81] Sasaki A, Hoen AG, Ozonoff A, et al. Evidence-based tool for triggering schoolclosures during influenza outbreaks, Japan. Emerg Infect Dis 2009;15:1841–3.

[82] Gostin L. Public health strategies for pandemic influenza: ethics and the law.JAMA 2006;295:1700–4.

Paediatric Respiratory Reviews 16 (2015) 127–132

Review

Paediatric lung recruitment: a review of the clinical evidence

Jacqui Jauncey-Cooke a,c,*, Chris E. East b, Fiona Bogossian a

a School of Nursing & Midwifery, The University of Queensland, Australiab School of Nursing and Midwifery/Maternity Services, Monash University/Southern Health, Clayton, Victoria and the School of Nursing & Midwifery,

The University of Queensland, Australiac Paediatric Critical Care Research Group, PICU, Mater Children’s Hospital, Brisbane, Australia

EDUCATIONAL AIMS

The reader will:

� Develop an understanding of the rationale for lung recruitment in mechanically ventilated paediatrics.� Learn the various means of achieving and measuring lung recruitment.� Be aware of the limitations of lung recruitment.

A R T I C L E I N F O

Keywords:

Paediatric

Mechanical ventilation

Lung recruitment

PEEP

Sustained inflation

VILI

S U M M A R Y

Lung recruitment is used as an adjunct to lung protective ventilation strategies. Lung recruitment is a

brief, deliberate elevation of transpulmonary pressures beyond what is achieved during tidal ventilation

levels. The aim of lung recruitment is to maximise the number of alveoli participating in gas exchange

particularly in distal and dependant regions of the lung. This may improve oxygenation and end

expiratory levels. Restoration of end expiratory levels and stabilisation of the alveoli may reduce the

incidence of ventilator induced lung injury (VILI). Various methods of lung recruitment have been

studied in adult and experimental populations. This review aims to establish the evidence for lung

recruitment in the pediatric population.

� 2014 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

INTRODUCTION

All mechanically ventilated paediatric patients are at risk ofdeveloping ventilator induced lung injury (VILI) and whilemechanical ventilation may be lifesaving, a proportion of thesechildren will acquire a degree of lung injury as a direct result of theventilation strategies employed by the clinician [1]. Mechanicalventilation strategies are continuously evolving but never more sothan following the American-European Consensus Conference,ARDSNet lung protective recommendations [2]. These lungprotective ventilation strategies (LPVS) were developed post large,multi-site randomised trials in adults [2,3]. Low tidal volumes,adequate PEEP, minimal peak pressures and minimal FiO2 are now

* Corresponding author. Tel.: +61 418335634(moble)/7 3163 1143 2(work).

E-mail addresses: jcqjaun@aol.com, j.jaunceycooke@uqconnect.edu.au,

jacqui.jauncey-cooke@mater.org.au (J. Jauncey-Cooke), eastc@unimelb.edu.au

(C.E. East), f.bogossian@uq.edu.au (F. Bogossian).

http://dx.doi.org/10.1016/j.prrv.2014.02.003

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

considered standard practice [2,3]. These recommendations havebeen universally extrapolated to paediatrics, possibly due to thepaucity of paediatric data. A retrospective cohort study byHalbertsma and colleagues in paediatrics did demonstrate acorrelation between high tidal volume ventilation and a reducedoxygenation ratio (P/F) which is consistent with the findings of theARDSNet study [4]. Ultimately, irrespective of how gently we treatthe paediatric lung, iatrogenic harm still occurs [1].

Lung recruitment is considered an adjunct to LPVS and in partaddresses the chronic derecruitment that occurs secondary to lowtidal volume ventilation [5,6]. This chronic derecruitment coupledwith the acute derecruitment that occurs with each circuitdisconnection contributes to the incidence of VILI. This reviewaims to assess the evidence for lung recruitment in mechanicallyventilated paediatrics.

A literature search was undertaken in 2014. Databases searchedincluded Medline, Embase, Lilac, Central register for Cochranereviews and Cinahl. The search yielded 70 abstracts. Following areview of the abstracts 8 relevant, paediatric papers were

Table 1Paediatric studies of lung recruitment

Subjects Method of recruitment Outcome measures Results

Boriosi et al. 2011 [15] Age yrs 4.8 (1-14)

ALI/ARDS

n=21

Established Cdyn

Incremental and decremental

PEEP via ventilator (Servo-I) in

Assist Control/Pressure Control

mode using Open Lung Tool

(OLT)

PaO2

PaCO2

-Improved P/F ratio for up to

12 hours post recruitment

-Nil adverse events

Boriosi et al.

2012 [16]

Age yrs 2.5 (0.5-14)

Nested study of ALI patients

from their 2011 study

n=6

Established Cdyn

Incremental and decremental

PEEP via

ventilator (Servo-I) in Assist

Control/Pressure Control mode

using Open Lung Tool (OLT)

End expiratory lung (EEL) levels

as measured by Computed

Tomography (CT) scan

Oxygenation

-Increase in EEL post RM 3%-

72% median 20% (IQR 6, 47)

-Reduction in PIP post RM by

-14% (IQR -18, -12)

-Improved P/F ratios

Duff et al.

2007 [17]

Age months 16

(11days to 14 years)

Lung disease status:

from healthy lungs through

to ARDS n=32

SI of 30-40cmH2O

for 15-20secs post

circuit interruption, suctioning,

hypoxia or routinely every

12hours

Oxygenation as

measured by P/F or S/F ratios

Haemodynamic markers

Safety

-Improved

oxygenation post

recruitment for up to 4 hours

-Safe in children

-Spikes in ICP in some

children

Halbertsma et al.

2010 [66]

Age months 0.5-

4.5

n= 7

Single recruitment manoeuvre –

incremental increases in PIP and

PEEP until transcutaneous SaO2

98%.

Max PEEP 30cmH2O Max PIP

45cmH2O

Translocation of

pulmonary cytokines

Oxygenation Lung kinetics

-Plasma levels of cytokines

increased post recruitment

-No increase inTcSaO2 noted

in 5/7 patients

-haemodynamic compromise

in 2/7 patients

Tusman

2003 [14]

Age months 6-72

Children undergoing

scheduled cranial MRI

n=24

Manual SI to

40cmH2O + PEEP of 15cmH2O for

10 breaths

Compared to 5 cm H2O of PEEP

alone and zero PEEP

% of atelectic regions as

measured by MRI

- Children in the recruitment

group had significantly less

atelectic regions compared to

those managed with PEEP

alone or zero PEEP

Marcus et al.

2002 [13]

Age < 24 months

Children undergoing

scheduled general

anaesthesia

n=20

Timed re-expansion inspiratory

manoeuvre =

30cmH2O CPAP for 10 seconds

Dynamic compliance

Airway resistance

-TRIM maneuver

resulted in improved

dynamic compliance

-Airway resistance changes

insignificant

Morrow et al.

2007 [36]

Age yrs <1

n=48

SI up to 30cmH2O for 30 seconds

using an anaesthetic bagging

circuit post ETT suctioning

Dynamic compliance

Oxygenation

Nil difference

between experimental and

control groups in terms of

oxygen saturations or

dynamic compliance

Wolf et al.

2012 [18]

Age yrs 9.9 � 4.2

ALI

n=10

Sustained inflation to 40cmH2O

for 40 secs using CPAP mode

followed by a stepwise RM,

escalating plateau pressures by

5cmH2O every 15mins

Regional atelectasis, lung

compliance and regional

overdistension as measured by

EIT

-small decrease in reversible

atelectasis post SI RM

-physiological lung

recruitment achieved in

responders during the

stepwise manoeuvre

-lung overdistension

proximally

J. Jauncey-Cooke et al. / Paediatric Respiratory Reviews 16 (2015) 127–132128

identified. The paediatric literature is referred to in the firstinstance. In the absence of paediatric literature, adult andexperimental literature is considered.

WHAT IS LUNG RECRUITMENT?

Lung recruitment is a deliberate strategy to increase transpul-monary pressure; to maximise the number of alveoli participatingin gas exchange [7]. The aim of lung recruitment is to recruit allrecruitable alveoli and minimize atelectic regions of the lung. It canbe achieved by either a sustained inflation (SI) or by brieflyincreasing positive end-expiratory pressure (PEEP). These methodsaim to overcome alveolar threshold opening pressures and/orovercome alveolar threshold closing pressures. Table 1 lists thecurrent evidence of lung recruitment in paediatrics.

WHOSE LUNGS DO YOU RECRUIT?

The process of lung recruitment is applicable to mechanicallyventilated children and those undergoing general anaesthesia.

Studies have been undertaken in both of these populations.Children, because of developmental differences may benefit mostfrom lung recruitment; physiologically immature lungs differsignificantly from adult lungs. At term, neonatal lungs possess only25% of their alveolar potential with a rapid increase in number inthe first two years of life and the interalveolar connections ofKohn’s pores are absent [8,9]. Diaphragmatic muscle fibre fatigue israpidly acquired as only 25% of muscle fibres are the fatigueresistant Type I – slow twitch fibres, compared to 50% at 8 monthsof age [9]. Additionally, the diaphragmatic angle is almosthorizontal rendering it less efficient in terms of contractility andoxidative capacity [9]. By approximately two years of age, chestwall and lung compliance is similar to adults however, olderinfants and children continue to have a significantly smaller airwayradius in proportion to their weight, less elastic retraction force,and a lower relaxation volume [10–12]. These factors all combineto predispose infants and children to atelectasis and hence anincreased risk of VILI. Subsequently these populations showpotential to benefit most from lung recruitment. In terms of lungdisease severity, which patients do you recruit? Using clinically

J. Jauncey-Cooke et al. / Paediatric Respiratory Reviews 16 (2015) 127–132 129

standard oxygenation ratios to crudely define the status of lungs;healthy lungs (PaO2/FiO2 {P/F} ratio) �300, acute lung injury (ALI)of 201-300 and acute respiratory distress syndrome (ARDS) �200;which of these cohorts benefit from lung recruitment? Two studieshave been undertaken on children with healthy lungs undergoingscheduled general anaesthesia [13,14]. Both of these studiesdemonstrated the benefit of lung recruitment in the healthy lung interms of improved compliance and a reduction in atelectasis[13,14]. A study by Boriosi and colleagues using incremental PEEPin children demonstrated an overall improvement in oxygenationthat was sustained for 12 hours in children with ALI/ARDS [15].Further subgroup analysis on the ALI population from this studydemonstrated an improvement in EEL and P/F ratio postrecruitment as well as a significant reduction in PIP [16]. A studyby Duff and colleagues in children with varied respiratory statusdemonstrated a significant and sustained reduction in FiO2

requirement for up to 6 hours post a SI manoeuvre [17]. Anotherstudy using SI as a means to recruit lungs in combination with astepwise recruitment manoeuvre reversed substantial atelectasisin dependant regions of the lung as measured by ElectricalImpedance Tomography (EIT) which was also associated withimprovements in gas exchange [18]. No recruitment studies havebeen undertaken exclusively in children with ARDS however astudy by Wolf and colleagues demonstrated the significant loss ofglobal and regional lung volume post endotracheal suctioning inchildren with ARDS as measured by EIT [19]. Potentially thosechildren with ARDS have the most to gain from lung recruitment.In adults, a number of studies have demonstrated a positive benefitof lung recruitment in patients with ARDS. A clinical study by Dyhrand colleagues used two hyperinflations to 45cmH2O, held for 20swith an interval of 60 seconds in patients with ARDS post-suctioning [20]. They found that the use of a recruitmentmanoeuvre, when combined with adequate ongoing PEEP,improved both oxygenation and EELV [20].

There is the potential that some children will be non-responders to lung recruitment and that the lung recruitmentmay in fact simply overdistend already recruited alveoli particu-larly in proximal regions. This was demonstrated in the study byWolf and colleagues as the reversal of atelectasis in dependantregions was accompanied by considerable overdistension innondependent lung regions [18]. This overdistension may con-tribute to barotrauma and subsequent VILI [18,19]. Children maybe more susceptible to overdistension than adults due to theircompliant chest wall [21,22]. In an experimental study, hyperin-flated regions of the lung shifted to normally aerated areas at PEEPlevels of 8cmH2O (mean of 30% decrease) and yet hyperinflatedareas still appeared at zero end expiratory pressure [23].Hyperinflation of easily distensible regions of the lung is anacknowledged limitation of recruitment manoeuvres [18,20,22–24].

HOW TO RECRUIT LUNGS?

Various methods of lung recruitment have been studied inpaediatrics but typically they involve either sustained inflations ormanipulating PEEP. A recent study involved incremental increasesin PEEP; an open lung strategy [15]. PEEP is used to maintainalveolar patency and stability at end-expiration but during tidalventilation is usually set at a level insufficient to overcome thethreshold opening pressure of atelectic alveoli [25–27]. IncreasingPEEP effectively allows the clinician to manipulate the functionalresidual capacity (FRC) and thereby minimise physiologicaldeadspace, improving ventilation/perfusion mismatch anddecreasing shunt fraction [6,12,28–30]. Boriosi and colleagueschose this method of recruitment to assess the safety and efficacyof the open lung tool in paediatrics. Ultimately they used a

modified version but found that oxygenation improved and peakinspiratory pressures were significantly reduced by 17% up to12 hours post recruitment [15,16]. Incremental increase in bothPeak Inspiratory Pressure (PIP) and PEEP has been shown to resultin improved gas exchange, tidal volume (Vt), EELV, oxygenationand lung compliance in ventilated adults [31–34].

Sustained inflations (SI) or inspiratory holds recruit alveoli via acombination of factors – the plateau pressure achieved and theduration of time maintained at that pressure. SI’s are reportedwithin the literature as ranging from 25-60cmH2O and held for aperiod of 10-30 seconds [14,17,18,20,35–37]. This is achieved byusing the Continuous Positive Airway Pressure (CPAP) setting onthe ventilator, the ‘inspiratory hold’ feature or manually by usinganaesthetic bagging circuits [17,18,34,36,37]. Irrespective oftechnique, the purpose of the SI is to maintain an elevated plateaupressure for a period of time greater than that achieved with tidalbreathing. A few studies have investigated the use of lungrecruitment in paediatrics using the SI method. The safety andefficacy of using a SI of 30-40cmH2O in paediatrics has beenestablished in one study when the SI was maintained for 15-20 seconds following circuit disconnection, suction, desaturationand/or routinely every 12 hours with positive results in terms ofoxygenation [17]. Another study in a PICU setting used a SI to30cmH2O for 30 seconds post endotracheal suctioning [36]. Thestudy design differed from similar studies in that there was a five-minute delay between suctioning and the lung recruitmentmanoeuvre and the SI was achieved using an anaesthetic bagwhich necessitated disconnection and then reconnection to theventilator circuit [36]. The authors concluded that the evidence didnot support recruitment manoeuvres; that dynamic compliancewas not improved post lung recruitment. The authors acknowl-edged that given the loss of EELV attributed to the disconnection ofthe circuit that any gains achieved from the recruitmentmanoeuvre could be reduced [36]. Another clinical study inhealthy children undergoing anaesthesia measured the extent ofatelectasis induced by an FiO2 of 1.0 and the rate of recruitmentwith a single SI procedure [13]. The results showed a significant re-recruitment of atelectatic alveoli at pressures of 25-30cmH2O [13].This study used what they refer to as a TRIM – a timed re-expansion inspiratory manoeuvre - involving a constant pressureof 30cmH2O for 10 seconds which resulted in significantly higherpulmonary compliance and lower airway resistance than thecontrol group.(P<0.0001) [13]. Other studies have also demon-strated significant improvements in compliance post recruitmentand in some instances oxygenation [33,34,37,38].

An alternative lung recruitment strategy is to use the inflexionpoints of the Pressure-Volume (P-V) curve derived from ventila-tory software. P-V curves can be used as a tool to manageventilation and recruitment as they describe the viscoelasticcharacteristics of the lung [13]. There is a general consensus in theliterature that opening of lung units occurs along the entire lengthof the inspiratory limb of the P-V relationship [37–40]. Theeffectiveness of PEEP in recruiting alveolar tissue correlates withthe value of the Lower Inflexion Point (LIP) [13,40,41]. A range ofventilators have software suitable to measure P-V curves. Amethod of determining an appropriate figure for SI was calculatedby Qiu and colleagues based upon 5 X Mean Airway Pressure(MAP). [42] This method provided optimal recruitment asmeasured by oxygenation, pulmonary dynamics, haemodynamicsand lung histology in an experimental study by Lindgren andcolleagures [42]. Some clinicians use the Upper Inflexion Points(UIP) and LIP’s to establish optimal PEEP and/or optimalrecruitment. Two studies have confirmed the effective use of PVloops to reflect recruitment above the LIP and below the UIP inadult populations [41,42]. However a study by Pestana andcolleagues found that P-V curves rarely reflected recruitment

J. Jauncey-Cooke et al. / Paediatric Respiratory Reviews 16 (2015) 127–132130

manoeuvres in ARDS patients which they suggest limits theirapplication in the clinical setting as a measure of recruitmentefficacy or as a guide to manage ventilation [43]. Two studiescaution clinicians against using inbuilt ventilator P-V software asthey had significant deleterious results in their research [44–47].P- V loops may provide clinicians with data regarding lung aerationyet do not distinguish between hyperinflation and recruitment andprovide no information in regard to distribution of ventilation [46].

Due to the heterogeneity of the techniques used to conduct lungrecruitment it is not possible to pool the data for further analysis.

WHEN DO YOU RECRUIT LUNGS?

Paediatric patients rely heavily upon their FRC to maintainairway and alveolar patency; this FRC steadily diminishes with theuse of LPVS [6,12]. During positive pressure ventilation there is aninhomogeneous distribution of gas, although the distribution ofalveolar opening pressures is deemed Gaussian [9]. Duff andcolleagues routinely conducted lung recruitment twice daily toaddress the progressive de-recruitment associated with LPVS [17].Airway pressure release, subsequent to circuit disconnection,results in a sudden and profound loss of FRC [36,42]. A number ofstudies have been conducted in adults demonstrating thesignificant losses associated with circuit disconnection and theapplication of suction to the airways [48–50]. One paediatric studycompared open and closed suction and the impact on lung volume[51]. Choong and colleagues demonstrated using inductiveplethysmography that the greatest loss of lung volume is relatedto disconnection from the circuit and subsequently recommendedthe use of closed suction units [51]. However a number of studieshave cautioned clinicians about negative pressures associated withthe use of closed suction units and a reduction in secretionclearance particularly in paediatrics [52–54]. Consequently, theuse of closed suction systems, at least in the paediatric population,remains questionable. In addition, saline instillation exacerbatesgas maldistribution and worsens de- recruitment by increasingthreshold opening pressures [51,52]. Recent studies measured theimpact of various suctioning methods and modes of ventilationand found that irrespective of ventilation mode, or whether openor closed suctioning was used, a significant loss of FRC occurred.[48–51] In a study by Lindgren et al, FRC as measured by EITdecreased by 58�24% of baseline after disconnection of theendotracheal tube and a further 22�10% during open suction itself[50]. It is probable that some form of recruitment manoeuvre may benecessary to attenuate the impact of LPVS, particularly after circuitdisconnection and the application of suction.

WHAT BENEFIT IS THERE TO USING LUNG RECRUITMENT?

Oxygenation

A number of studies have explored the impact of lungrecruitment on oxygenation in the paediatric population. Boriosiand colleagues noted a P/F ratio increase of 53% (p< .01) in childrenpost lung recruitment that was sustained for up to 12 hours[15,16]. Duff and colleagues found a reduced oxygen demand forup to 6 hours post a sustained inflation lung recruitmentmanoeuvre [17]. In adults several studies have demonstrated abeneficial effect on oxygenation assuming adequate baseline PEEPis present [7,20,34,44,55].

Lung volume

Both CT and EIT have been used in studies to measure lungvolume post recruitment in paediatrics. A study by Wolf andcolleagues uses EIT to measure the impact of a combined SI and

incremental PEEP manoeuvre on end expiratory lung volume inventilated children diagnosed with ALI [18]. In their analysis theyclearly identified responders and non-responders to lung recruit-ment [18]. The children that responded to the recruitment werecharacterised by resolution of atelectasis in the most dependantregions of the lungs by 17% � 4% (p = .016) plus an improvement incompliance across all regions of the lung [18]. The non-respondershad no discernable reduction in atelectasis in the most dependantregions nor did they have a significant change in compliance [18].Overdistension was problematic with the incremental PEEP man-oeuvre in both the responders and non-responders [18].

Three studies measuring lung volume in intubated paediatricshave been performed in anaethetised children with healthy lungs.One such study demonstrated a reduction in atelectic regionswhen a PEEP of 5cmH2O was applied compared to zero PEEP [56].Another study measured the impact of various inspiratorypressures on atelectic regions on children undergoing a CT scan[35]. They found that those children ventilated with higherpressures (30cmH20) had significantly less atelectasis than thoseventilated to pressures of 25cmH2O [35]. A number of adult studieshave measured lung volume using either computed tomography(CT) scans or EIT. A clinical study by Dyhr and colleagues used twohyperinflations to 45cmH2O, held for 20s with an interval of 60s inpatients with ARDS post-suctioning [20]. They found that the useof a recruitment manoeuvre, when combined with adequateongoing PEEP, improved EELV [20]. Another study by Gattinoniused CT to measure the positive gains of increased PEEP in reducingatelectic regions and that these gains correlated with oxygenationincreases [57].

Lung water

The ALI and ARDS lung is typically oedematous [58,59]. Nostudies have measured the impact of lung recruitment on totallung water volume in children however a study by Toth andcolleagues in adults did conclude that improvements in oxygena-tion post recruitment was independent of redistribution ofextravascular lung water [60]. Another study demonstrated anet alveolar fluid clearance post recruitment which the authorsattributed to resorption of alveolar oedema [61].

Compliance

Two studies have measured the impact of lung recruitment oncompliance. Morrow and colleagues found no improvement indynamic compliance with their lung recruitment technique usingan open ended anaesthetic circuit post endotracheal suctioning ininfants [36]. Boriosi and colleagues also found no significantchange in compliance [15,16]. However, Marcus and colleaguesfound a 30% increase in dynamic pulmonary compliance usingVentrak software post their TRIM manouvre (p <0.01) [13].

Morbidity

No studies to date have measured the short or long term impactof lung recruitment on morbidity in paediatrics.

Mortality

No studies to date have measured the impact of lungrecruitment on mortality in paediatrics.

WHAT HARM CAN LUNG RECRUITMENT CAUSE?

Lung recruitment is considered by some clinicians to becontroversial. Concerns include the potential for barotrauma,

J. Jauncey-Cooke et al. / Paediatric Respiratory Reviews 16 (2015) 127–132 131

pneumothoraces and the possible contributory factor to both VILIand ventilator associated pneumonia (VAP) [12,20,36,44,60]. Otherconcerns include the potential to impact negatively on haemody-namics and intracranial pressure [17]. By increasing the intrathor-acic pressure during the recruitment manoeuvre an inevitableconsequence is a commiserate reduction in venous return andsubsequent cardiac output. Additionally, overdistension of alveoliwill increase the regional pulmonary vascular resistance andsubsequently will decrease regional perfusion [13]. The method ofrecruitment and the pressures used obviously impacts on thedegree of any deleterious impact. An experimental study byOdenstedt and colleagues measured the haemodynamic and lungmechanic side effects of three different recruitment manoeuvres; avital capacity maneuver to 40cmH2O, a pressure controlledmaneuver up to 40cmH2O with a PEEP of 20cmH2O and a slowrecruitment with PEEP to 15cmH2O held for 7 seconds [62]. Theyfound that the slow, lower pressure manoeuvre produced lessnegative lung mechanic side effects and less circulatory depression[62]. The higher pressure manoeuvres produced the greatest lungexpansion but the slower, lower pressure manoeuvre produced thegreatest improvement in oxygenation [62]. Transient reduction incardiac output have been noted in experimental studies [39,62–65]. There is the potential that lung recruitment may contribute tothe barotrauma associated with mechanical ventilation. Anexperimental study by Frank and colleagues determined thedegree of alveolar and lung endothelial injury post a sustainedinflation recruitment manoeuvre [39]. While oxygenation andcompliance improved with the recruitment manoeuvre those inthe recruitment group did appear to have a protected endotheliumbut there was no reduction in alveolar epithelial injury [39]. Onestudy investigating the potential of harm of lung recruitment hasbeen conducted in paediatrics. Halbertsma and colleagues foundthat a single sustained inflation recruitment manoeuvre withinspiratory pressures up to �45cmH2O and PEEP pressures of�30cmH2O translocated cytokines into the circulation [66].Translocation of bacteria has been discovered in two studies usingpressures in excess of 45cmH2O [67,68]. An adult studyinvestigated the potential of lung recruitment manoeuvres toreduce gastric mucosal perfusion [69]. Measuring gastric mucosalperfusion with laser Doppler flowmetry they found that nosignificant reduction in perfusion occurred during lung recruit-ment [69]. Another study produced a marked yet transitoryimpairment in splanchnic circulation for up to 8 minutes post-manoeuvre [70].

CONCLUSION

Recruitment manoeuvres may be useful in both restoring lungvolume post circuit disconnection +/- endotracheal suctioning butalso to minimise the chronic derecruitment subsequent to LPVSalthough insufficient evidence exists to support it’s universalapplication. The various methods of recruiting lungs have as theircommon goal the prompt restoration of EELV to improverespiratory compliance and oxygenation and to minimise alveolarshearing and subsequent VILI. Recruitment manoeuvres appearmost effective when applied following disconnection of the ETTfrom the circuit and following airway suctioning. Various methodsof lung recruitment have been tested. The current evidence doesnot enable us to reach a consensus as to which method is mosteffective, nor which patients are most receptive to lung recruit-ment manoeuvres. Haphazard application of recruitment man-oeuvres may cause harm and as such should be individuallyassessed for each patient. Whether the consistent use ofrecruitment manoeuvres will reduce morbidity and mortalityassociated with mechanical ventilation is yet to be determined.Further research on the efficacy of various recruitment methods

and the timing of recruitment in paediatrics is required to informour clinical practice.

FUTURE RESEARCH DIRECTIONS

� Establish the optimal method of lung recruitment given thedegree of lung injury.� Establish definitive clinical markers of ‘responders’ to lung

recruitment.� Measure the effect of lung recruitment on morbidity and

mortality.

References

[1] Dahlem P, van Aalderen W, Hamaker M, Kijkgraaf M, Bos A. Incidence andshort-term outcome of acute lung injury in mechanically ventilated children.European Respiratory Journal 2003;22:980–5.

[2] The ARDS Network. Ventilation with lower tidal volumes as compared withtraditional tidal volumes for acute lung injury and the acute respiratorydistress syndrome. The New England Journal of Medicine 2000;342:1301–8.

[3] Artigas A, Bernard G, Carlet J, Dreyfuss D, Gattanoni L, Hudson L, et al. TheAmerican-European Consensus Conference on ARDS, Part 2 Ventilatory, Phar-macologic, Supportive Therapy. Study design strategies and issues related torecovery and remodelling. Intensive Care Medicine 1998;24:378–98.

[4] Halbertsma F, Vaneker M, Pickkers P, Hoeven J. The oxygenation ratio duringmechanical ventilation in children: The role of tidal volume and positive end-expiratory pressure. Journal of Critical Care 2009;24:220–6.

[5] Moran I, Zavala E, Fernandez R, Blanch L, Mancebo J. Recruitment manoeuvresin acute lung injury/acute respiratory distress syndrome. European RespiratoryJournal 2003;22. S42-37s-42s.

[6] Barbas C, et al. Mechanical ventilation in acute respiratory failure: Recruitmentand high positive end-expiratory pressure are necessary. Current Opinion inCritical Care 2005;11(1):18–28.

[7] Lapinsky S. Recruitment and retention of lung volume. Critical Care2003;7(1):9–10.

[8] Priestley M, Helfaer M. Approaches in the management of acute respiratoryfailure in children. Current Opinion in Pediatrics 2004;16(3):293–8.

[9] Taussig L, Landau L. Pediatric Respiratory Medicine. Le Souef: Mosby; 1999.[10] Papastamelos C, Panitch H, England S, Allen L. Developmental changes in chest

wall compliance in infancy and early childhood. Journal Applied Physiology1995;78:179–84.

[11] Suki B. Fluctuations and Power Laws in Pulmonary Physiology. AmericanJournal Respiratory Critical Care Medicine 2002;166:133–7.

[12] von Ungern-Sternberg B, Regli A, Schibler A, Hammer J, Frei F, Erb T. The Impactof Positive End-Expiratory Pressure on Functional Residual Capacity andVentilation Homogeneity Impairment in Anesthetised Children Exposed toHigh Levels of Inspired Oxygen. Anesthesia Analgesia 2007;104:1364–8.

[13] Marcus R, van der Walt J, Pettifer R. Pulmonary volume recruitment restorespulmonary compliance and resistance in anaesthetized young children. Pae-diatric Anaesthesia 2002;12:579–84.

[14] Tusman G, et al. Effects of Recruitment Maneuver on atelectasis in anesthe-tized children. Anesthesiology 2003;98:14–22.

[15] Boriosi J, et al. Efficacy and safety of lung recruitment in pediatric patients withacute lung injury. Pediatric Critical Care Medicine 2011;12(4):431–6.

[16] Borisosi J, et al. Lung aeration changes after lung recruitment in children withacute lung injury. Pediatric Pulmonology 2012;47:771–9.

[17] Duff J, Rosychuk R, Joffe A. The safety and efficacy of sustained inflations as alung recruitment maneuver in pediatric intensive care unit patients. IntensiveCare Medicine 2007;33:1778–86.

[18] Wolf G, et al. Reversal of dependent lung collapse predicts response to lungrecruitment in children with early acute lung injury. Pediatric Critical CareMedicine 2012;13(5):509–15.

[19] Wolf G, et al. Regional lung volume changes in children with acute respiratorydistress syndrome during a derecruitment manoeuvre. Critical Care Medicine2007;35(8):1972–8.

[20] Dyhr T, Bonde J, Larsson A. Lung recruitment manoeuvres are effective inregaining lung volume and oxygenation after open endotracheal suctioning inacute respiratory distress syndrome. Critical Care 2002;7:55–62.

[21] Neve V, Leclerc F, Dumas de la Roque E, Leteurtre S, Riou Y. Overdistension inventilated children. Critical Care 2001;5(4):196–203.

[22] Papastamelos C, Panitch H, England S, Allen J. Developmental changes in chestwall compliance in infancy and early childhood. Journal of Applied Physiology1995;78(1):179–84.

[23] Carvalho A, Jandre F, Pino A, Bozza F, Salluh J, Rodrigues R, et al. Positive end-expiratory pressure at minimal respiratory elastance represents the bestcompromise between mechanical stress and lung aeration in oleic acidinduced lung injury. Critical Care 2007;11(R86).

[24] Grasso S, et al. Effects of recruiting maneuvers in patients with acute respira-tory distress syndrome ventilated with protective ventilatory strategy. Anaes-thesiology 2002;96:795–802.

J. Jauncey-Cooke et al. / Paediatric Respiratory Reviews 16 (2015) 127–132132

[25] The ARDS Network. Effects of recruitment maneuvers in patients with acutelung injury and acute respiratory distress syndrome ventilated with highpositive end-expiratory pressure. Critical Care Medicine 2003;31(11):2592–7.

[26] Halter J, et al. Effect of positive end-expiratory pressure and tidal volume onlung injury induced by alveolar instability. Critical Care 2007;11(1).

[27] Lim S-C, et al. Intercomparison of recruitment maneuver efficacy in threemodels of acute lung injury. Critical Care Medicine 2004;32(12):2371–7.

[28] Gattinoni L, D’Andrea L, Pelosi P, Vitale G, Pesenti A, Fumagalli R. Regionaleffects and mechanism of positive end-expiratory pressure in early adultrespiratory distress syndrome. Journal of American Medical Association1993;269(16):2122–7.

[29] Papadakos P, Lachmann B. The Open Lung Concept of Mechanical Ventilation:The role of Recruitment and Stabilization. Critical Care Clinics 2007;23(2):241–50.

[30] Rimensberger P, Pache J, mcDerlie C, Frndova H, Cox P. Lung recruitment andlung volume maintenance: a strategy for improving oxygenation and pre-venting lung injury during both conventional mechanical ventilation andhigh-frequency oscillation. Intensive Care Medicine 2000;26:745–55.

[31] Tusman G, Bohm S, Suarezz-Sipmanet F, Turchetto E. Lung recruitmentimproves the efficiency of ventilation and gas exchange during one-lungventilation anesthesia. Anesthesia Analgesia 2004;98:1604–9.

[32] Povoa P, Almeida E, Fernandes A, Mealha R, Moreira P, Sabino H. Evaluation ofa recruitment maneuver with positive inspiratory pressure and high PEEP inpatients with severe ARDS. Acta Anaesthesiology Scandinavica 2004;48(3):287–93.

[33] Borges J, Okomoto V, Matos G, Caramez M, Arantas P, Barros F, et al. Rever-sibility of lung collapse and Hypoxemia in Early Acute Respiratory DistressSyndrome. American Journal Respiratory Critical Care Medicine 2006;174:268–78.

[34] Dyhr T, Nygard e, Laursen N, Larsson A. Both lung recruitment maneuver andPEEP are needed to increase oxygenation and lung volume after cardiacsurgery. Acta Anaesthesiology Scandinavica 2004;48(2):187–97.

[35] Sargent M, Jamieson D, McEachern A, Blackstock D. Increased inspiratorypressure for reduction of atelectasis in children anesthetized for CT scan.Pediatrics Radiology 2002;32:344–7.

[36] Morrow B, Futter M, Argent A. A recruitment manoeuvre performed afterendotracheal suction does not increase dynamic compliance in ventilatedpaediatric patients: a randomised controlled trial. Australian Journal of Phy-siotherapy 2007;53:163–9.

[37] Tugrul S, et al. Effects of sustained inflation and postinflation positive end-expiratory pressure in acute respiratory distress syndrome: focusing onpulmonary and extrapulmonary forms. Critical Care Medicine 2003;31:974–6.

[38] Almgren B, Wickerts C-J, Hogman M. Post-suction Recruitment ManoeuvreRestores Lung Function in Healthy, Anaesthetized Pigs. Anaesthesia and Inten-sive Care 2004;32(3):339–45.

[39] Frank J, Matthay M. Mechanisms of ventilator-induced injury. Critical Care2002;7:233–41.

[40] Henzler D, Hochhausen N, Dembinski R, Orfao S, Rossaint R, Kuhlen R. Para-meters derived from the pulmonary pressure-volume curve, but not thepressure-time curve, indicate recruitment in experimental lung injury. Anaes-thesia & Analgesia 2007;105(4):1072–8.

[41] Rothen H, Neumann P, Berglund JE, Valtysson J, Magnusson A, Hedenstierna G.Dynamics of re-expansion of atelectasis during general anaesthesia. BritishJournal of Anaesthesia 1999;82(4):551–7.

[42] Lindgren S, Odenstedt H, Olegard C, Sondergaard S, Lundin S, Stenqvsit O.Regional lung derecruitment after endotracheal suction during volume- orpressure-controlled ventilation: a study using electric impedance tomogra-phy. Intensive Care Medicine 2007;33(1):172–80.

[43] Pestana D, Hernandez-Gancedo, Royo C, Perez-Chrzanowska Criado A. Pres-sure-volume curve variations after a recruitment manoeuvre in acute lunginjury/ARDS patients: implications for the understanding of the inflectionpoints of the curve. European Journal of Anaesthesiology 2005;22:175–80.

[44] Lapinsky S, Aubin M, Mehta S, Boiteau P, Slutsky A. Safety and efficacy of asustained inflation for alveolar recruitment in adults with respiratory failure.Intensive Care Medicine 1999;25(11):1297–301.

[45] Syring R, Otto C, Spivack R, Markstaller K, Baumgardner J. Maintenance of end-expiratory recruitment with increased respiratory rate after saline-lavagelung injury. Journal Applied Physiology 2007;102:331–9.

[46] Maggiore S, Richard J-C, Brochard L. What has been learnt from P/V curves inpatients with acute lung injury/acute respiratory distress syndrome? Eur-opean Respiratory Journal 2003;22(Suppl. 42):22s–6s.

[47] Henzler D, Mahnken A, Dembinski r, Waskowiak B, rossaint R, Kuhlen R.Repeated generation of the pulmonary pressure-volume curve may lead toderecruitment in experimental lung injury. Intensive Care Medicine2005;31(2):302–10.

[48] Heinze H, Sedemund-Adib B, Herinlake M, Gosch U, Eichler W. FunctionalResidual Capacity changes after different endotracheal suctioning methods.Anesthesia Analgesia 2008;107(3):941–4.

[49] Cereda M, Villa F, Colombo E, Greco G, Nacoto M, Pesenti A. Closed systemendotracheal suctioning maintains lung volume during volume-controlledmechanical ventilation. Intensive Care Medicine 2001;27:648–54.

[50] Lindgren S, et al. Effectiveness and side effects of closed and open suctioning:an experimental evaluation. Intensive Care Medicine 2004;30:1630–7.

[51] Choong K, Chatrkaw P, Frndova H, Cox P. Comparison of loss in lung volumewith open versus in-line catheter endotracheal suctioning. Paediatric CriticalCare Medicine 2003;4:69–73.

[52] Stenqvist O, Lindgren S, Karason S, Sondergaard S, Lundin S. Warning! Suc-tioning. A lung model evaluation of closed suctioning systems. Acta Anaes-thesiology Scandinavica 2001;45:167–72.

[53] Copnell B, et al. A comparison of the effectiveness of open and closedendotracheal suction. Intensive Care Medicine 2007;33(9):1655–62.

[54] Kiraly N, Tingay D, Mills F, Morley C, Copnell B. Negative tracheal pressureduring neonatal endotracheal suction. Pediatric Research 2008;64:29–33.

[55] Schreiter D, et al. Alveolar recruitment in combination with sufficient positiveend-expiratory pressure increases oxygenation and lung aeration in patientswith severe chest trauma. Critical Care Medicine 2004;32(4):968–75.

[56] Serafini G, et al. Pulmonary atelectasis during paediatric anaesthesia: CT scanevaluation and effect of positive endexpiratory pressure (PEEP). PediatricAnesthesia 1999;9(3):225–8.

[57] Gattinoni L, et al. Relationships between lung computed tomographic density,gas exchange, and PEEP in acute respiratory failure. Anesthesiology 1988;69:824–32.

[58] Sibbard W, et al. Thermal dye measurements of extravascular lung water incritically ill patients. Intravascular Starling forces and extravascular lungwater in the adult respiratory distress syndrome. Chest 1985;87:585–92.

[59] Kirov M, et al. Extravascular lung water assessed by transpulmonary singlethermodilution and post-mortem gravimetry in sheep. Critical Care2004;8:R451–8.

[60] Toth I, et al. Hemodynamic and respiratory changes during lung recruitmentand descending optimal positive end-expiratory pressure titration in patientswith Acute Respiratory Distress Syndrome. Critical Care Medicine 2007;35(3):787–93.

[61] Constantin J-M, et al. Response to recruitment maneuver influences netalveolar fluid clearance in acute respiratory distress syndrome. Anesthesiology2007;106(5):944–51.

[62] Odenstedt H, et al. Slow moderate pressure recruitment maneuver minimizesnegative circulatory and lung mechanic side effects: evaluation of recruitmentmaneuvers using electric impedance tomography. Intensive Care Medicine2005;31:1706–14.

[63] Hansen L, et al. Selective recruitment maneuvers for lobar atelectasis: effectson lung function and central hemodynamics: An experimental study in pigs.Anesthesia and Analgesia 2006;102:1504–10.

[64] Lim S-C, et al. Transient hemodynamic effects of recruitment maneuvers inthree experimental models of acute lung injury. Critical Care Medicine 2004;32(12):2378–84.

[65] Nielsen J, et al. Central hemodynamics during lung recruitment maneuvers athypovolemia, normovolemia and hypervolemia. A study by echocardiographyand continuous pulmonary artery flow measurements in lung-injured pigs.Intensive Care Medicine 2006;32:585–94.

[66] Halbertsma F, et al. A single recruitment maneuver in ventilated critically illchildren can translocate pulmonary cytokines into the circulation. Journal ofCritical Care 2010;25:10–5.

[67] Verbrugge S, et al. Lung overinflation without positive end-expiratory pres-sure promotes bacteremia after experimental Klebsiella pneumonieae inocu-lation. Intensive Care Medicine 1998;24:172–7.

[68] Tremblay L, Valenza F, Ribeiro S, Li J, Slutsky S. Injurious ventilatory strategiesincrease cytokines and c-fos m-RNA expression in an isolated rat lung model.Journal Clinical Investigation 1997;99:944–52.

[69] Claesson J, Lehtipalo S, Winso O. Do lung recruitment maneuvers decreasegastric mucosal perfusion? Intensive Care Medicine 2003;29(3):1314–21.

[70] Nunes S, Rothen H, Brander L, Takala J, Jakob S. Changes in splanchniccirculation during an alveolar recruitment maneuver in healthy porcine lungs.Anesthesia Analgesia 2004;98:1432–8.

Paediatric Respiratory Reviews 16 (2015) 133–139

Review

Risk and Protective Factors for the Development of Childhood Asthma

Guodong Ding 1,2, Ruoxu Ji 1, Yixiao Bao 1,*1 Department of Pediatrics, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China2 MOE and Shanghai Key Laboratory of Children’s Environment Health, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China

EDUCATIONAL AIMS

1. Asthma, as a complex disease, has a broad spectrum of potential determinants ranging from genetics to environmental andlifestyle-related factors.

2. Despite evidence indicating that many factors are probably associated with the onset of childhood asthma, the relationships arenot considered causal.

3. Only environmental tobacco smoke has been associated with an increased risk for the development of childhood asthma.

A R T I C L E I N F O

Keywords:

Childhood asthma

Development

Risk factor

Protective factor

Environment

Lifestyle

S U M M A R Y

Childhood asthma prevalence worldwide has been increasing markedly over several decades. Various

theories have been proposed to account for this alarming trend. The disease has a broad spectrum of

potential determinants ranging from genetics to lifestyle and environmental factors. Epidemiological

observations have demonstrated that several important lifestyle and environmental factors including

obesity, urban living, dietary patterns such as food low in antioxidants and fast food, non-breastfeeding,

gut flora imbalance, cigarette smoking, air pollution, and viral infection are associated with asthma

exacerbations in children. However, only environmental tobacco smoke has been associated with the

development of asthma. Despite epidemiological studies indicating that many other factors are probably

associated with the development of asthma, the relationships are not considered causal due to the

inadequate evidence and inconsistent results from recent studies. This may highlight that sufficient data

and exact mechanisms of causality are still in need of further study.

� 2014 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Paediatric Respiratory Reviews

INTRODUCTION

Over recent decades, the prevalence of childhood asthma hasbeen dramatically increasing globally, but the etiology of thedisease is still not well understood. Asthma incidence among U.S.children increased from 3.6% in 1980 to 5.8% in 2003, and it is thethird highest cause of hospitalization, exceeded only by pneumo-nia and injuries [1]. Increases in the prevalence of asthma withsimilar or even greater magnitude were also reported from

* Corresponding author. Department of Pediatrics, Xinhua Hospital, Shanghai Jiao

Tong University School of Medicine, 1665 Kongjiang Road, 200092 Shanghai, China.

Tel.: +86-21-25078300.

E-mail address: drsmilebao@163.com (Y. Bao).

http://dx.doi.org/10.1016/j.prrv.2014.07.004

1526-0542/� 2014 Elsevier Ltd. All rights reserved.

many other developed countries such as the U.K., Germany,Canada, and Australia [1]. Although asthma is generally lesscommon in developing countries than in developed countries, theprevalence is increasing as they become more westernized orcommunities become urbanized. Such a transition is currentlytaking place in China at a much higher speed and during a shorterperiod than in many other countries [2]. The second nationwidesurvey in 2000 revealed that the prevalence of asthma amongChinese children 0�14 years old was 1.97%, nearly 2 times of thatin 1990 (1.00%), suggesting an increasing trend [3]. Asthma, as acomplex disease, has a broad spectrum of potential determinantsranging from genetics to environmental and lifestyle factors.However, this rapidly increasing incidence worldwide cannot beexplained by genetic causes alone, as genetic changes requiremany generations for population-wide effects to occur. Evidence

G. Ding et al. / Paediatric Respiratory Reviews 16 (2015) 133–139134

has shown that environmental and lifestyle factors are likely to bekey events in explaining the overall increasing trend towardasthma prevalence [4]. In this report, we aimed to provide the datarelating several important lifestyle and environmental factors, andto discuss their possible association with the development ofchildhood asthma.

Lifestyle Factors

Overweight/obesity

Recently, the prevalence of both asthma and obesity in childrenhave increased substantially in many countries. The parallel rise inprevalence of both disorders and the coexistence of both asthma andobesity has led interest in the association between the two epidemics[5]. A number of epidemiological studies in children have examinedthe relationship of obesity or overweight with asthma, and manystudies support a positive association between body mass index (BMI)and asthma. Chinn performed a comprehensive review and found thatobesity increases the risk of subsequent asthma, although no evidencesupports the hypothesis that asthma leads to increased obesity [6].Flaherman and Rutherford conducted a meta-analysis and found thathigh birth weight had a pooled relative risk of 1.2 for the subsequentdevelopment of asthma, and further calculated a populationattributable risk of 0.066 [7].

It should be noted that children and adolescents or boys andgirls were considered together in many previous studies which didnot divide children on age or gender. One problem exists here thatdoes not exist in the adult studies, which is in the definition ofweight for height in children and, more particularly, in the cut-offpoints used to define overweight and obesity [8]. In children, adultBMI cut-off points are not an accurate measure of body fatnessbecause BMI changes with age, requiring age-specific cut-offpoints [5]. Unfortunately, until now there has been no standard-ized definition of overweight and obesity in children to correspondwith the adult cut-off points. The other problem is that sex may bean important confounding factor in the study of obesity andasthma. However, the data is conflicting as to whether theassociation of BMI with asthma is affected by gender. A prospectivestudy revealed that among overweight children, the risk of new-onset asthma was evident among boys but not in girls [9]. Incontrast, another birth cohort study showed that, girls, but notboys, who became overweight were seven times more likely todevelop asthma [10].

Urban/rural area

Many epidemiological studies in the past decade haveconsistently documented that children living in rural areas seemto be at lower risk of asthma than their urban counterparts;moreover, children living in a rural areas are at decreased risk ofdeveloping asthma [11]. For example, rural Chinese children hadsignificantly lower prevalence of asthma and atopic sensitizationthan urban children, using the validated ISAAC (InternationalStudy of Asthma and Allergies in Childhood) questionnaire andobjective skin-prick tests [12]. Another large study in the U.S.pediatric Medicaid population found that the rural children hadincreased asthma prevalence and similar asthma morbiditycompared with urban children [13]. These results support thehygiene theory, early exposure to infection for children may conferan advantage by regulating the immune system to protect againstallergies so as to reduce the future risk of asthma.

Although the underlying mechanisms behind this apparentprotective effect of rural/farm living are not well understood, theoverall consensus is that environmental factors and socioeconomicissues predispose people to asthma. It appears that places thatshare similar environmental and socioeconomic risk factors mayhave comparable prevalence of asthma regardless of whether it is

in a rural or urban location [11]. These factors may partly explainwhy there were no differences in asthma prevalence in rural andurban areas from several studies [14,15]. However, not all farmingenvironments are associated with protection against allergies. Themulticenter PARSIFAL study revealed that exposure to sheepfarming was associated with increased risk of allergies, but theexplanation for this association remained to be explored [16].

Diet (antioxidants)

The antioxidant hypothesis was first proposed in 1994 bySeaton et al., who suggested that alteration in diet associated withwesternization may be responsible for the increase in asthmaprevalence [17]. Observations showed that consumption of foodsrich in antioxidants had decreased in the UK diet while asthmaprevalence concurrently rose. The transition from a traditional dietto a modern diet appeared to have resulted in a decrease inantioxidant intake [17]. Subsequently, many observational studieshave focused on vitamin C, vitamin E, carotenoids, flavonoids, andminerals such as selenium and zinc, and typically these havereported low antioxidant intake to be associated with an increasedincidence of childhood asthma [18,19]. However, not all studies onthe role of antioxidants have been positive. A meta-analysisconcluded that dietary intake of antioxidant vitamins C and E andb–carotene does not significantly influence the risk of asthma [20].In addition, the potential role of antioxidants as supplements hasbeen explored, but a number of studies have been inconclusive. ACochrane review of vitamin C supplementation in asthma showedthat there is insufficient evidence to recommend vitamin Csupplementation in the treatment of asthma [21]. It should benoted that overall the body of observational evidence is inherentlyweak because of the biases and limitations of cross-sectional andcase-control studies that predominate. These limitations includethe difficulties in quantifying dietary intake, reverse causation, andlack a temporal element. Unfortunately, longitudinal data are verylimited for antioxidants highlighted in studies of asthma. There isan urgent need for longitudinal studies to fill the gaps ofinformation on the association of antioxidants with asthma.

Diet (lipids)

Black and Sharpe in 1997 proposed that the rise of asthmaprevalence may stem from increased consumption of polyunsatu-rated fatty acids (PUFAs) and decreased consumption of saturatedfat [22]. The v-6 PUFAs may particularly have a role in regulatingimmune response and inflammation. These PUFAs are foundlargely as linoleic acid in foods such as margarine and vegetableoils, which have increased in consumption with westernization[23]. Linoleic acid is a precursor of prostaglandin E2 that inhibitsinterferon-g and promotes an inflammatory environment whichfavors the development of asthma. Meanwhile, v-3 PUFAs fromoily fish may have an anti-inflammatory role [23]. Therefore, atopicsensitization and inflammation could be promoted by increasingdietary intake of v-6 PUFAs and decreasing intake of v-3 PUFAs.

A small number of epidemiological studies have examined thelipid hypothesis, but they reported inconsistent results. Forexample, in children aged 12 to 15 years, atopic disease andatopic sensitization were associated with reduced v-3 PUFA and anincrease in v-6/v-3 PUFA ratio. In addition, serum IgE levels werepositively associated with v-6 PUFAs and negatively associatedwith serum eicosapentaenoic acids (EPAs) [24]. In contrast, Grieseet al reported that plasma and mononuclear cell phospholipid EPAlevels were positively associated with atopic asthma and serum IgEin children [25]. Although there is increasing interest in the use ofdietary PUFA supplementation to prevent the development ofasthma and atopic disease, it is disappointing that interventionstudies have not found consistent results nor provided sufficientsupport for dietary supplementation with PUFAs [26,27].

G. Ding et al. / Paediatric Respiratory Reviews 16 (2015) 133–139 135

Early hypotheses analyzed population level trends and focusedon major dietary factors such as antioxidants and lipids. Morerecently, larger dietary patterns beyond individual nutrients havebeen investigated such as fast foods and the Mediterranean diet[23]. A cross-sectional study of children aged 10 to 12 years in NewZealand revealed that hamburger consumption positively associ-ated with asthma symptoms while takeaway consumption had amarginal effect on bronchial hyperresponsiveness [28]. On theother hand, the Mediterranean diet has been suggested as ahealthy dietary pattern that may reduce the risk of asthma [29].Despite the plethora of cross-sectional data about fast foods andthe Mediterranean diet, there is a lack of longitudinal studies andanalyses to form a causal link between these foods and asthmaprevalence.

Breastfeeding

Many studies have demonstrated that breastfeeding, inaddition to its nutritional and sociological benefits, provided anumber of specific health benefits to the infant, includingreduction of the incidence of allergy and childhood asthma [30].A number of studies reported lower risks of asthma, atopic eczema,and positive allergy skin tests in breastfed children, or equivalent-ly, higher risks in infants fed conventional cow milk- or soy-basedformulas [31], and many of these studies reported a greater degreeof protection with more exclusive and/or more prolongedbreastfeeding [32,33]. The third NHANES study in the U.S. showedthat compared with never breastfed children, ever breastfedchildren had significantly reduced odds of being diagnosed withasthma and of having recurrent wheeze before 24 months of age[34]. A multidisciplinary review from the Swedish NationalInstitute of Public Health concluded that exclusive breastfeedingreduces the risk of developing asthma, and the protective effectsincrease with the duration of the breastfeeding up to at least 4months of age [35].

Yet, several publications have challenged this view, particu-larly with respect to the long-term outcomes for asthma.Breastfeeding does not necessarily protect children againstasthma and the magnitude of the effect is relatively modest,and may even increase the risk [36,37]. Sears et al. conducted alongitudinal study of children from New Zealand, and found anincreased risk of asthma in children who were breastfed [38]. Atotal of 1037 children were recruited at birth and assessed atregular intervals from 9 years of age for the presence of asthmaand atopy. Breastfeeding significantly increased the likelihood ofcurrent asthma at age 9 years (11% vs 5%), at 15 years (18% vs 11%),at 21 years (19% vs 13%), and at 26 years (23% vs 15%) [38]. Theseconclusions have been preeminent in raising concerns about theeffects of breastfeeding on the development of asthma, and manyscholars have raised several critical questions and challengedthese conclusions [39]. First, the breastfeeding data werecollected at age 3 years, which may indicate problems with theaccuracy of recall. Second, individuals were classified as beingbreastfed if they had some breastfeeding for at least 4 weeks butmost cohort studies classified as being breastfed had more than 12weeks breastfeeding. Third, breastfeeding was not exclusive, assupplementation with formula at night was common in NewZealand. Evidence has shown that a beneficial effect of breast-feeding on allergic disease indicating that both the duration, andexclusivity are of importance.

To date no plausible mechanism explains how breastfeedingcould increase the risk of asthma. On the other hand, evidence hasshown major benefits of breastfeeding for child neurodevelop-ment and chronic disease (e.g., obesity). It is thus important toconsider that any decision not to breastfeed made in the light ofpossible adverse effects on the development of asthma is not wellfounded.

Probiotics

The human gastrointestinal tract is sterile at birth, rapidlyundergoing colonization of the gut with subsequent developmentof the immune system. Variations in patterns of microbialcolonization of the gastrointestinal tract, linked with lifestyleand geographic factors, may be important determinants of theheterogeneity in allergy prevalence throughout the world. Studieshave shown that there are obvious differences in the compositionof intestinal microbiota between healthy and allergic infantswithin the first week of life and before clinical symptoms for thelatter group, suggesting that modifying microbiota compositionmay affect disease outcome [40].

Probiotics are increasingly marketed and used as healthpromoting agents, which mainly contain beneficial bacteria suchas Lactobacilli and bifidobacteria. Several previous studies sug-gested that probiotics used as dietary supplements may beeffective in preventing early atopy in children through themodulation of intestinal microbiota and the regulation ofinflammatory response [41,42]. A recent cohort study shows thatprobiotic milk consumption in pregnancy is associated with aslightly reduced incidence of eczema and rhinoconjunctivitis, butnot asthma, at 3 years of age [43]. Despite these findings indicatinga beneficial effect of lactic acid-producing bacteria, manyquestions have been raised regarding optimal strains, dosages,timing and route of administration. Ongoing clinical trials arerequired before any recommendations can be given about the useof probiotics in the prevention or treatment of allergy.

Environmental Exposure

Environmental tobacco smoke/active smoking

A number of epidemiological surveys support the role ofenvironmental tobacco smoke (ETS) exposure in increasing theincidence of wheezing, airway hyperresponsiveness, and asthma inchildren [44,45]. In a population-based cohort study in Finland thatincluded almost 60,000 children, the risk of developing asthmaamongst children at 7 years of age increased in a dose-dependentmanner with maternal smoking rates during pregnancy: OR 1.23(95% CI 1 .07–1.42) for <10 cigarettes/day and OR 1.35 (95% CI1.13–1.62) for >10 cigarettes/day [44]. Consequently, a series ofsystematic reviews and meta-analyses provide compelling evi-dence of a causal relationship between parental smoking and thedevelopment of childhood asthma [46,47]. For example, Burkeet al. identified over 70 studies examining the effect of exposure topre- or post-natal passive smoke on the development of asthma,and reported a 21% to 85% increased risk of incident asthma. Thestrongest effect from prenatal maternal smoking on asthma inchildren aged � 2 years had an OR of 1.85 (95% CI 1.3–2.5) [47]. Itseems difficult to separate the effect of prenatal ETS exposure fromthe effect of postnatal ETS exposure because women who smokedduring pregnancy are likely to continue smoking after delivery.Still, there is some evidence to indicate that the associationbetween prenatal exposure to maternal smoking with wheeze and/or asthma in children is stronger than that of postnatal exposure[48].

Among adolescents with asthma, exposure to tobacco smoke,not only by ETS but also by use of tobacco smoke products (activesmoking), can cause increased asthma morbidity includingreduced lung function and exacerbation of asthma symptoms[49,50]. Current asthma treatment guidelines recommend avoid-ing exposure to tobacco smoke, including both ETS and activesmoking. However, most of previous studies have focused on activesmoking and risk of developing asthma in adulthood. It is debatedwhether active smoking induces the development of asthma inadolescents. Few studies performed in adolescents have shownthat active smoking increased the risk of new-onset of asthma

G. Ding et al. / Paediatric Respiratory Reviews 16 (2015) 133–139136

during adolescence. In a prospective study including 2609 childrenwith no lifetime history of asthma, Gilliland et al. found thatchildren who were nonsmokers without any history of allergy andwho became regular smokers later in life were 5.2 times morelikely to develop asthma, but among those with a history of allergythere was little evidence of increased risk [51].

Building upon existing findings, exposure to passive smokingincreases the risk of incident asthma in children. There is an needfor more evidence to confirm whether active smoking inadolescents could increase the risk of developing asthma.

Indoor air pollution

Major categories of indoor pollutants include nitrogen dioxide(NO2), particulate matter (PM), and indoor allergens (e.g., dustmite, mouse, cockroach, and molds). NO2, a common air pollutant,is produced from high temperature combustion. Cooking with gasis by far the most important source of indoor NO2 levels in Chinaand around the world [52]. Historically, several epidemiologicalstudies evaluated the potential health effects of indoor NO2

exposure on childhood asthma by using the gas appliances as asurrogate for indoor NO2 levels, and many of these studies reportedthat the use of gas appliances was associated with an increased riskof asthma [53,54]. Recently, population-based studies withmeasured indoor NO2 and risk of developing asthma in childrenhave emerged but produced inconsistent results, with somereporting a positive relationship [55] and others showing noassociation [56]. Despite recent indoor monitoring studies thathave been performed to develop better exposure estimates forNO2, these approaches to quantifying exposure have severelimitations. Daily or weekly average exposures to NO2 weremeasured or estimated without error, but short term peakexposures were not adequately known. Indeed, the major sourcesof NO2 indoors are typically in use for only a few hours each day,and NO2 levels follow a spiked pattern with one hour maximumlevels exceeding daily or weekly averages [57]. Animal studiessuggested that repeated exposure to high NO2 levels for shortperiods of time is more harmful than continuous exposure toelevated NO2 levels at a lower level [58]. To sum up, population-based studies in children observed less consistent results on theassociation between asthma and exposures to NO2, assessed by thepresence of gas appliances or indoor nitrogen dioxide measure-ments.

PM, consists of coarse PM10 and fine PM2.5, is a principalcomponent of indoor air pollution. Indoor sources include cookingexhaust, wood-burning stoves and fireplaces, cleaning activities,and penetration of outdoor particles. Indoor PM differs fromoutdoor PM in source, composition, and concentration, and thehealth effects of indoor PM cannot be readily extrapolated fromstudies of outdoor air pollution. Previous studies and meta-analysis have indicated that exposure to high indoor (both coarseand fine) PM levels is associated with decreased lung function andrespiratory symptoms in children with asthma [59,60]. However,few studies regarding indoor PM and the onset of asthma inchildren are available, and thus a causal link has not beenestablished.

Common indoor allergens include house dust mite, cockroach,animal dander, and certain molds. Evidence from epidemiologicalstudies indicates that exposure to indoor allergen in sensitizedindividuals with asthma has clear implications on lung functionand asthma severity [61,62]. The counterargument that reductionin indoor allergen exposure will result in improvement in asthmahas been less convincing [63]. Although cross-sectional studiessupported the hypothesis that allergen exposure causes asthma[64,65], the evidence is considered weak as confounder bias cannotbe totally ruled out. Prospective studies may be better designedregarding the role of indoor allergens in the causation of asthma,

but many of them reported inconsistent results. In a large birthcohort from Germany, Lau et al. showed that no association wasfound between early indoor allergen exposure and the develop-ment of asthma or bronchial hyperresponsiveness in childrenobserved up to 7 years of age [66]. In contrast, a recent study fromthe same cohort found that children who were sensitized and hadhigh exposure to the relevant allergen were at high risk ofpersistent asthma and bronchial hyperresponsiveness in laterchildhood [67]. Therefore, evidence for the direct associationbetween exposure to indoor allergens and the development ofchildhood asthma is available but not conclusive.

Ambient air pollution

With rapid urbanization in many communities, traffic (diesel)exhausts have become the major source of ambient/outdoor airpollution. Therefore, there have been a number of studies indicatingthat exposure to ambient air pollutants can exacerbate pre-existingasthma in children [68,69]. However, there are relatively few studiesthat address the issue of whether traffic exhausts may induce thedevelopment of asthma. Earlier studies have generally relied onsimple measures of traffic proximity and density to estimateexposure and have not shown an association between air pollutionand asthma incidence [70,71]. More recent studies have usedmodeling approaches that provide high-resolution estimates ofneighborhood-scale variations in air pollution. Several studies usingthis approach have observed increases in asthma incidence forchildren exposed to higher levels of traffic-related air pollution[72,73]. However, not all such studies of this type have reportedconsistent associations [74]. For example, as part of an internationalcollaborative research on the impact of Traffic-Related Air Pollutionon Childhood Asthma, two cohort studies from Holland and Swedenshowed a positive association [72,73], but another cohort study fromGermany showed no association [74]. Thus, it may be accepted thatambient air pollution can exacerbate asthma in those who alreadyhave the condition. Whether air pollution can contribute to thedevelopment of asthma remains speculative due to the existingevidence for causal relationship being unconvincing.

Viral infection

Viral respiratory infections, particularly with respiratorysyncytial virus (RSV) and human rhinovirus (HRV), are the mostcommon causes of wheezing in infants and young children and arecommon triggers of asthma exacerbations in pediatric patientswith preexisting asthma [75,76]. Whether these infections arecausal in asthma development or simply identify predisposedchildren remains a controversial issue. Contrary to the concept thatlink infections in early life with subsequent wheezing, the hygienehypothesis proposed that respiratory infections in early life wereprotective towards the development of asthma. However, thishypothesis has not been well supported by evidence, such as anincrease of asthma in North American inner cities that aregenerally characterized by poor housing and dirty environments[77]. More recent studies indicate that exposure to non-pathogenicrather than pathogenic microbes would be more important inreducing the risk for asthma [78].

RSV lower respiratory tract illnesses in infancy, particularlythose severe enough to lead to hospitalization, are associated withsubsequent recurrent episodes of wheezing [79,80]. This associa-tion has prompted speculation that RSV lower respiratory illnessmay be causal in asthma development. In this regard, Sigurs et al.conducted a case-control study to examine the associationbetween an RSV infection and the eventual development ofasthma. They found that severe RSV bronchiolitis was associatedwith an increased risk of asthma at 13 years of age [81]. A largeretrospective cohort study in Tennessee supported a causal role forRSV bronchiolitis during infancy on asthma inception [82].

G. Ding et al. / Paediatric Respiratory Reviews 16 (2015) 133–139 137

However, an association between RSV infection and asthma hasnot been seen in all investigations. A longitudinal data from theTucson Children’s Respiratory Study suggested that RSV infectionsof the lower respiratory tract during the first 3 years of life wereassociated with subsequent wheezing and asthma in earlychildhood, but not beyond age 11 years [83]. Similarly, Thomsenet al. examined the relationship between severe RSV infection anddevelopment of asthma based on a registry-based twin study inDenmark, and found that severe RSV infection (enough to warranthospitalization) does not cause asthma but is an indicator of thegenetic predisposition to asthma [84].

With the development of molecular diagnostics, HumanRhinovirus (HRV) wheezing illnesses have been recognized morerecently as a stronger predictor of developing asthma than RSV[85]. Several epidemiological observations have identified thatinfants hospitalized with HRV wheezing illnesses have a increasedrisk of developing asthma later in childhood. Kotaniemi-Syrjanenet al. conducted a long-term case-control study and found thatHRV-induced wheezing episodes in infancy were highly predictiveof subsequent asthma, and this relationship persisted at leastthrough the late teen years [86]. This finding was subsequentlyconfirmed by the results of one birth cohort study. The ChildhoodOrigins of Asthma (COAST) study, a high-risk birth cohortexamining the role of respiratory viruses in development ofasthma, identified HRV wheezing illnesses during the first year oflife as significant risk factors for wheezing in the third year of life[87] and asthma at age six years [88].

Collectively, early respiratory RSV infections such as RSV andHRV seem to be an important risk factor associated with short-term recurrent wheeze, but whether RSV or HRV is causal inasthma inception is an open question. Despite existing evidenceindicating that wheezing with HRV may be the most robustpredictor of subsequent asthma, adequate data of causality are stillin need of further study.

Finally, it should be noted that asthma is a complex diseaseresulting from genetic and environmental interactions andepigenetic regulation is also a major contributor. It is well knownthat many individuals are predisposed to developing allergicreactions to substances that do not in general elicit an immuneresponse, and these individuals are thought to be geneticallypredisposed to develop hypersensitivity to substances such aspollens and perfumes [89]. On the other hand, the environmentalcomponent is also critical, as supported by many studiesdocumenting shifting asthma rates by geographical region andlevel of urbanization, and others documenting differences indisease incidence between monozygotic twins [90]. More andmore emerging evidence suggests that epigenetic regulationfollowing environmental exposures may underlie the interfacebetween prenatal and early postnatal environmental exposure andthe development of asthma [90]. Epigenetic changes can occurthroughout life, but much of the epigenome is established duringearly development of the fetus. Several prenatal environmentalexposures such as maternal smoking, dietary pattern, andmicrobial exposure have been shown to modify fetal immunefunction important to the later development of asthma and otherallergic diseases [91–93]. How genes and environment exactlyinteract and then contribute to the development of allergicdiseases is yet poorly understood, and the role of epigeneticregulation in asthma susceptibility is a new and promising areathat needs further attention and better understanding.

CONCLUSIONS

Numerous risk and protective factors have been identifiedwhich could influence the likelihood of sensitization and triggersymptoms in already sensitized individuals. However, only one

environmental factor (passive smoking) has been associated withan increased risk for the development of childhood asthma.Despite evidence indicates that many other factors are probablyassociated with the onset of asthma, the relationships are notconsidered causal due to the inadequate evidence and inconsistentresults from current epidemiological studies. Childhood asthmaprobably has complex etiologies involving the interaction betweenthe environment, genetic susceptibility, and chance (critical timewindows). Equipped with this general understanding, a logical andessential step is to examine these factors simultaneously in anattempt to further elucidate the existing causal relationships forchildhood asthma. As the development of asthma is probablymultifactorial, there is a need for more large-scale studies that arecapable of providing sufficient power to detect even modestassociations with precision.

References

[1] Eder W, Ege MJ, von Mutius E. The asthma epidemic. N Engl J Med2006;355(21):2226–35.

[2] Yang G, Kong L, Zhao W, Wan X, Zhai Y, Chen LC, Koplan JP. Emergence ofchronic non-communicable diseases in China. Lancet 2008;372(9650):1697–705.

[3] National cooperation group on childhood asthma. A nationwide survey inChina on prevalence of asthma in urban children [in Chinese]. Zhonghua Er KeZa Zhi 2003;41(2):123–7.

[4] Gern JE. The Urban Environment and Childhood Asthma study. J Allergy ClinImmunol 2010;125(3):545–9.

[5] Story RE. Asthma and obesity in children. Curr Opin Pediatr 2007;19(6):680–4.[6] Chinn S. Obesity and asthma. Paediatr Respir Rev 2006;7(3):223–8.[7] Flaherman V, Rutherford GW. A meta-analysis of the effect of high weight on

asthma. Arch Dis Child 2006;91(4):334–9.[8] Chinn S. Obesity and asthma: evidence for and against a causal relation. J

Asthma 2003;40(1):1–16.[9] Gilliland FD, Berhane K, Islam T, McConnell R, Gauderman WJ, Gilliland SS,

Avol E, Peters JM. Obesity and the risk of newly diagnosed asthma in school-age children. Am J Epidemiol 2003;158(8):406–15.

[10] Castro-Rodrıguez JA, Holberg CJ, Morgan WJ, Wright AL, Martinez FD. In-creased incidence of asthmalike symptoms in girls who become overweight orobese during the school years. Am J Respir Crit Care Med 2001;163(6):1344–9.

[11] Malik HU, Kumar K, Frieri M. Minimal difference in the prevalence of asthma inthe urban and rural environment. Clin Med Insights Pediatr 2012;6:33–9.

[12] Ma Y, Zhao J, Han ZR, Chen Y, Leung TF, Wong GW. Very low prevalence ofasthma and allergies in schoolchildren from rural Beijing, China. PediatrPulmonol 2009;44(8):793–9.

[13] Valet RS, Gebretsadik T, Carroll KN, Wu P, Dupont WD, Mitchel EF, Hartert TV.High asthma prevalence and increased morbidity among rural children in aMedicaid cohort. Ann Allergy Asthma Immunol 2011;106(6):467–73.

[14] Priftis KN, Anthracopoulos MB, Nikolaou-Papanagiotou A, Matziou V, PaliatsosAG, Tzavelas G, Nicolaidou P, Mantzouranis E. Increased sensitization in urbanvs. rural environment–rural protection or an urban living effect? PediatrAllergy Immunol 2007;18(3):209–16.

[15] Pesek RD, Vargas PA, Halterman JS, Jones SM, McCracken A, Perry TT. Acomparison of asthma prevalence and morbidity between rural and urbanschoolchildren in Arkansas. Ann Allergy Asthma Immunol 2010;104(2):125–31.

[16] Ege MJ, Frei R, Bieli C, Schram-Bijkerk D, Waser M, Benz MR, Weiss G, Nyberg F,van Hage M, Pershagen G, Brunekreef B, Riedler J, Lauener R, Braun-FahrlanderC, von Mutius E, PARSIFAL Study team. Not all farming environments protectagainst the development of asthma and wheeze in children. J Allergy ClinImmunol 2007;119(5):1140–7.

[17] Seaton A, Godden DJ, Brown K. Increase in asthma: a more toxic environmentor a more susceptible population? Thorax 1994;49(2):171–4.

[18] Harik-Khan RI, Muller DC, Wise RA. Serum vitamin levels and the risk ofasthma in children. Am J Epidemiol 2004;159(4):351–7.

[19] Rubin RN, Navon L, Cassano PA. Relationship of serum antioxidants to asthmaprevalence in youth. Am J Respir Crit Care Med 2004;169(3):393–8.

[20] Gao J, Gao X, Li W, Zhu Y, Thompson PJ. Observational studies on the effect ofdietary antioxidants on asthma: a meta-analysis. Respirology 2008;13(4):528–36.

[21] Kaur B, Rowe BH, Arnold E. Vitamin C supplementation for asthma. CochraneDatabase Syst Rev 2009;1:CD000993.

[22] Black PN, Sharpe S. Dietary fat and asthma: is there a connection? Eur Respir J1997;10(1):6–12.

[23] Kim JH, Ellwood PE, Asher MI. Diet and asthma: looking back, moving forward.Respir Res 2009;10:49.

[24] Yu G, Bjorksten B. Polyunsaturated fatty acids in school children in relation toallergy and serum IgE levels. Pediatr Allergy Immunol 1998;9(3):133–8.

[25] Griese M, Schur N, Laryea MD, Bremer HJ, Reinhardt D, Biggemann B. Fatty acidcomposition of phospholipids of plasma and of mononuclear blood cells inchildren with allergic asthma and the influence of glucocorticoids. Eur J Pediatr1990;149(7):508–12.

G. Ding et al. / Paediatric Respiratory Reviews 16 (2015) 133–139138

[26] Dunstan JA, Mori TA, Barden A, Beilin LJ, Taylor AL, Holt PG, Prescott SL. Fish oilsupplementation in pregnancy modifies neonatal allergen-specific immuneresponses and clinical outcomes in infants at high risk of atopy: a randomized,controlled trial. J Allergy Clin Immunol 2003;112(6):1178–84.

[27] Almqvist C, Garden F, Xuan W, Mihrshahi S, Leeder SR, Oddy W, Webb K, MarksGB, CAPS team. Omega-3 and omega-6 fatty acid exposure from early life doesnot affect atopy and asthma at age 5 years. J Allergy Clin Immunol2007;119(6):1438–44.

[28] Wickens K, Barry D, Friezema A, Rhodius R, Bone N, Purdie G, Crane J. Fast foods- are they a risk factor for asthma? Allergy 2005;60(12):1537–41.

[29] Chatzi L, Apostolaki G, Bibakis I, Skypala I, Bibaki-Liakou V, Tzanakis N,Kogevinas M, Cullinan P. Protective effect of fruits, vegetables and the Medi-terranean diet on asthma and allergies among children in Crete. Thorax2007;62(8):677–83.

[30] Duncan JM, Sears MR. Breastfeeding and allergies: time for a change inparadigm? Curr Opin Allergy Clin Immunol 2008;8(5):398–405.

[31] Kramer MS. Breastfeeding and allergy: the evidence. Ann Nutr Metab2011;59(Suppl 1):20–6.

[32] Kull I, Almqvist C, Lilja G, Pershagen G, Wickman M. Breast-feeding reduces therisk of asthma during the first 4 years of life. J Allergy Clin Immunol2004;114(4):755–60.

[33] Snijders BE, Thijs C, Dagnelie PC, Stelma FF, Mommers M, Kummeling I,Penders J, van Ree R, van den Brandt PA. Breast-feeding duration and infantatopic manifestations, by maternal allergic status, in the first 2 years of life(KOALA study). J Pediatr 2007;151(4):347–51.

[34] Chulada PC, Arbes Jr SJ, Dunson D, Zeldin DC. Breast-feeding and the preva-lence of asthma and wheeze in children: analyses from the Third NationalHealth and Nutrition Examination Survey, 1988–1994. J Allergy Clin Immunol2003;111(2):328–36.

[35] van Odijk J, Kull I, Borres MP, Brandtzaeg P, Edberg U, Hanson LA, Høst A,Kuitunen M, Olsen SF, Skerfving S, Sundell J, Wille S. Breastfeeding and allergicdisease: a multidisciplinary review of the literature (1966–2001) on the modeof early feeding in infancy and its impact on later atopic manifestations. Allergy2003;58(9):833–43.

[36] Wright AL, Holberg CJ, Taussig LM, Martinez FD. Factors influencing therelation of infant feeding to asthma and recurrent wheeze in childhood. Thorax2001;56(3):192–7.

[37] Oddy WH, Peat JK, de Klerk NH. Maternal asthma, infant feeding, and the riskof asthma in childhood. J Allergy Clin Immunol 2002;110(1):65–7.

[38] Sears MR, Greene JM, Willan AR, Taylor DR, Flannery EM, Cowan JO, HerbisonGP, Poulton R. Long-term relation between breastfeeding and development ofatopy and asthma in children and young adults: a longitudinal study. Lancet2002;360(9337):901–7.

[39] Kemp A, Kakakios A. Asthma prevention: breast is best? J Paediatr Child Health2004;40(7):337–9.

[40] Bjorksten B. Effects of intestinal microflora and the environment on thedevelopment of asthma and allergy. Springer Semin Immunopathol2004;25(3–4):257–70.

[41] Kalliomaki M, Salminen S, Arvilommi H, Kero P, Koskinen P, Isolauri E.Probiotics in primary prevention of atopic disease: a randomised placebo-controlled trial. Lancet 2001;357(9262):1076–9.

[42] Rosenfeldt V, Benfeldt E, Nielsen SD, Michaelsen KF, Jeppesen DL, Valerius NH,Paerregaard A. Effect of probiotic Lactobacillus strains in children with atopicdermatitis. J Allergy Clin Immunol 2003;111(2):389–95.

[43] Bertelsen RJ, Brantsæter AL, Magnus MC, Haugen M, Myhre R, Jacobsson B,Longnecker MP, Meltzer HM, London SJ. Probiotic milk consumption inpregnancy and infancy and subsequent childhood allergic diseases. J AllergyClin Immunol 2014;133(1):165–71.

[44] Jaakkola JJ, Gissler M. Maternal smoking in pregnancy, fetal development, andchildhood asthma. Am J Public Health 2004;94(1):136–40.

[45] Goksor E, Amark M, Alm B, Gustafsson PM, Wennergren G. The impact of pre-and post-natal smoke exposure on future asthma and bronchial hyper-re-sponsiveness. Acta Paediatr 2007;96(7):1030–5.

[46] Vork KL, Broadwin RL, Blaisdell RJ. Developing asthma in childhood fromexposure to secondhand tobacco smoke: insights from a meta-regression.Environ Health Perspect 2007;115(10):1394–400.

[47] Burke H, Leonardi-Bee J, Hashim A, Pine-Abata H, Chen Y, Cook DG, Britton JR,McKeever TM. Prenatal and passive smoke exposure and incidence of asthmaand wheeze: systematic review and meta-analysis. Pediatrics2012;129(4):735–44.

[48] Gilliland FD, Li YF, Peters JM. Effects of maternal smoking during pregnancyand environmental tobacco smoke on asthma and wheezing in children. Am JRespir Crit Care Med 2001;163(2):429–36.

[49] Mannino DM, Homa DM, Redd SC. Involuntary smoking and asthma severity inchildren: data from the Third National Health and Nutrition ExaminationSurvey. Chest 2002;122(2):409–15.

[50] Yoo S, Kim HB, Lee SY, Kim BS, Kim JH, Yu J, Kim BJ, Lee DH, Seong MW, Hong SJ.Effect of active smoking on asthma symptoms, pulmonary function, and BHRin adolescents. Pediatr Pulmonol 2009;44(10):954–61.

[51] Gilliland FD, Islam T, Berhane K, Gauderman WJ, McConnell R, Avol E, PetersJM. Regular smoking and asthma incidence in adolescents. Am J Respir Crit CareMed 2006;174(10):1094–100.

[52] Heinrich J. Influence of indoor factors in dwellings on the development ofchildhood asthma. Int J Hyg Environ Health 2011;214(1):1–25.

[53] Dekker C, Dales R, Bartlett S, Brunekreef B, Zwanenburg H. Childhood asthmaand the indoor environment. Chest 1991;100(4):922–6.

[54] Ponsonby AL, Couper D, Dwyer T, Carmichael A, Kemp A, Cochrane J. Therelation between infant indoor environment and subsequent asthma. Epide-miology 2000;11(2):128–35.

[55] van Strien RT, Gent JF, Belanger K, Triche E, Bracken MB, Leaderer BP. Exposureto NO2 and nitrous acid and respiratory symptoms in the first year of life.Epidemiology 2004;15(4):471–8.

[56] Garrett MH, Hooper MA, Hooper BM, Abramson MJ. Respiratory symptoms inchildren and indoor exposure to nitrogen dioxide and gas stoves. Am J RespirCrit Care Med 1998;158(3):891–5.

[57] Brunekreef B, Houthuijs D, Dijkstra L, Boleij JS. Indoor nitrogen dioxideexposure and children’s pulmonary function. J Air Waste Manage Assoc1990;40(9):1252–6.

[58] World Health Organization (WHO). Air quality guidelines for Europe (SecondEdition). Regional Office for Europe. Copenhagen. WHO Regional Publications,European Series, 2000; No. 91.

[59] McConnell R, Berhane K, Gilliland F, Molitor J, Thomas D, Lurmann F, Avol E,Gauderman WJ, Peters JM. Prospective study of air pollution and bronchiticsymptoms in children with asthma. Am J Respir Crit Care Med2003;168(7):790–7.

[60] Weinmayr G, Romeo E, De Sario M, Weiland SK, Forastiere F. Short-term effectsof PM10 and NO2 on respiratory health among children with asthma orasthma-like symptoms: a systematic review and meta-analysis. Environ HealthPerspect 2010;118(4):449–57.

[61] Arshad SH, Hamilton RG, Adkinson Jr NF. Repeated aerosol exposure to smalldoses of allergen. A model for chronic allergic asthma. Am J Respir Crit Care Med1998;157(6 Pt 1):1900–6.

[62] Salo PM, Arbes Jr SJ, Crockett PW, Thorne PS, Cohn RD, Zeldin DC. Exposure tomultiple indoor allergens in US homes and its relationship to asthma. J AllergyClin Immunol 2008;121(3):678–84.

[63] Morgan WJ, Crain EF, Gruchalla RS, O’Connor GT, Kattan M, Evans 3rd R, StoutJ, Malindzak G, Smartt E, Plaut M, Walter M, Vaughn B, Mitchell H, Inner-CityAsthma Study Group. Results of a home-based environmental interventionamong urban children with asthma. N Engl J Med 2004;351(11):1068–80.

[64] Nowak D, Heinrich J, Jorres R, Wassmer G, Berger J, Beck E, Boczor S, ClaussenM, Wichmann HE, Magnussen H. Prevalence of respiratory symptoms, bron-chial hyperresponsiveness and atopy among adults: west and east Germany.Eur Respir J 1996;9(12):2541–52.

[65] Dharmage S, Bailey M, Raven J, Mitakakis T, Cheng A, Guest D, Rolland J, ForbesA, Thien F, Abramson M, Walters EH. Current indoor allergen levels of fungiand cats, but not house dust mites, influence allergy and asthma in adults withhigh dust mite exposure. Am J Respir Crit Care Med 2001;164(1):65–71.

[66] Lau S, Illi S, Sommerfeld C, Niggemann B, Bergmann R, von Mutius E, Wahn U.Early exposure to house-dust mite and cat allergens and development ofchildhood asthma: a cohort study. Multicentre Allergy Study Group. Lancet2000;356(9239):1392–7.

[67] Illi S, von Mutius E, Lau S, Niggemann B, Gruber C, Wahn U, Multicentre AllergyStudy (MAS) group. Perennial allergen sensitisation early in life and chronicasthma in children: a birth cohort study. Lancet 2006;368(9537):763–70.

[68] Nicolai T, Carr D, Weiland SK, Duhme H, von Ehrenstein O, Wagner C, vonMutius E. Urban traffic and pollutant exposure related to respiratory outcomesand atopy in a large sample of children. Eur Respir J 2003;21(6):956–63.

[69] Zhao Z, Zhang Z, Wang Z, Ferm M, Liang Y, Norback D. Asthmatic symptomsamong pupils in relation to winter indoor and outdoor air pollution in schoolsin Taiyuan, China. Environ Health Perspect 2008;116(1):90–7.

[70] Wjst M, Reitmeir P, Dold S, Wulff A, Nicolai T, von Loeffelholz-Colberg EF, vonMutius E. Road traffic and adverse effects on respiratory health in children. BMJ1993;307(6904):596–600.

[71] English P, Neutra R, Scalf R, Sullivan M, Waller L, Zhu L. Examining associationsbetween childhood asthma and traffic flow using a geographic informationsystem. Environ Health Perspect 1999;107(9):761–7.

[72] Brauer M, Hoek G, Smit HA, de Jongste JC, Gerritsen J, Postma DS, Kerkhof M,Brunekreef B. Air pollution and development of asthma, allergy and infectionsin a birth cohort. Eur Respir J 2007;29(5):879–88.

[73] Nordling E, Berglind N, Melen E, Emenius G, Hallberg J, Nyberg F, Pershagen G,Svartengren M, Wickman M, Bellander T. Traffic-related air pollution andchildhood respiratory symptoms, function and allergies. Epidemiology2008;19(3):401–8.

[74] Gehring U, Cyrys J, Sedlmeir G, Brunekreef B, Bellander T, Fischer P, Bauer CP,Reinhardt D, Wichmann HE, Heinrich J. Traffic-related air pollution andrespiratory health during the first 2 yrs of life. Eur Respir J 2002;19(4):690–8.

[75] MacDowell AL, Bacharier LB. Infectious triggers of asthma. Immunol Allergy ClinNorth Am 2005;25(1):45–66.

[76] Carroll KN, Hartert TV. The impact of respiratory viral infection on wheezingillnesses and asthma exacerbations. Immunol Allergy Clin North Am2008;28(3):539–61.

[77] Platts-Mills TA, Erwin E, Heymann P, Woodfolk J. Is the hygiene hypothesis stilla viable explanation for the increased prevalence of asthma? Allergy2005;60(Suppl 79):25–31.

[78] Busse WW, Lemanske Jr RF, Gern JE. Role of viral respiratory infections inasthma and asthma exacerbations. Lancet 2010;376(9743):826–34.

[79] Kusel MM, de Klerk NH, Kebadze T, Vohma V, Holt PG, Johnston SL, Sly PD.Early-life respiratory viral infections, atopic sensitization, and risk of subse-quent development of persistent asthma. J Allergy Clin Immunol2007;119(5):1105–10.

[80] Hall CB, Weinberg GA, Iwane MK, Blumkin AK, Edwards KM, Staat MA, AuingerP, Griffin MR, Poehling KA, Erdman D, Grijalva CG, Zhu Y, Szilagyi P. The burden

G. Ding et al. / Paediatric Respiratory Reviews 16 (2015) 133–139 139

of respiratory syncytial virus infection in young children. N Engl J Med2009;360(6):588–98.

[81] Sigurs N, Gustafsson PM, Bjarnason R, Lundberg F, Schmidt S, Sigurbergsson F,Kjellman B. Severe respiratory syncytial virus bronchiolitis in infancy andasthma and allergy at age 13. Am J Respir Crit Care Med 2005;171(2):137–41.

[82] Wu P, Dupont WD, Griffin MR, Carroll KN, Mitchel EF, Gebretsadik T, HartertTV. Evidence of a causal role of winter virus infection during infancy in earlychildhood asthma. Am J Respir Crit Care Med 2008;178(11):1123–9.

[83] Stein RT, Sherrill D, Morgan WJ, Holberg CJ, Halonen M, Taussig LM, Wright AL,Martinez FD. Respiratory syncytial virus in early life and risk of wheeze andallergy by age 13 years. Lancet 1999;354(9178):541–5.

[84] Thomsen SF, van der Sluis S, Stensballe LG, Posthuma D, Skytthe A, Kyvik KO,Duffy DL, Backer V, Bisgaard H. Exploring the association between severerespiratory syncytial virus infection and asthma: a registry-based twin study.Am J Respir Crit Care Med 2009;179(12):1091–7.

[85] Jackson DJ, Lemanske Jr RF. The role of respiratory virus infections in childhoodasthma inception. Immunol Allergy Clin North Am 2010;30(4):513–22.

[86] Kotaniemi-Syrjanen A, Vainionpaa R, Reijonen TM, Waris M, Korhonen K,Korppi M. Rhinovirus-induced wheezing in infancy–the first sign of childhoodasthma? J Allergy Clin Immunol 2003;111(1):66–71.

[87] Lemanske Jr RF, Jackson DJ, Gangnon RE, Evans MD, Li Z, Shult PA, Kirk CJ, ReisdorfE, Roberg KA, Anderson EL, Carlson-Dakes KT, Adler KJ, Gilbertson-White S,

Pappas TE, Dasilva DF, Tisler CJ, Gern JE. Rhinovirus illnesses during infancypredict subsequent childhood wheezing. J Allergy Clin Immunol 2005;116(3):571–7.

[88] Jackson DJ, Gangnon RE, Evans MD, Roberg KA, Anderson EL, Pappas TE, PrintzMC, Lee WM, Shult PA, Reisdorf E, Carlson-Dakes KT, Salazar LP, DaSilva DF,Tisler CJ, Gern JE, Lemanske Jr RF. Wheezing rhinovirus illnesses in early lifepredict asthma development in high-risk children. Am J Respir Crit Care Med2008;178(7):667–72.

[89] Mukherjee AB, Zhang Z. Allergic asthma: influence of genetic and environ-mental factors. J Biol Chem 2011;286(38):32883–9.

[90] de Planell-Saguer M, Lovinsky-Desir S, Miller RL. Epigenetic regulation: theinterface between prenatal and early-life exposure and asthma susceptibility.Environ Mol Mutagen 2014;55(3):231–43.

[91] Noakes PS, Holt PG, Prescott SL. Maternal smoking in pregnancy alters neona-tal cytokine responses. Allergy 2003;58(10):1053–8.

[92] Chatzi L, Torrent M, Romieu I, Garcia-Esteban R, Ferrer C, Vioque J, KogevinasM, Sunyer J. Mediterranean diet in pregnancy is protective for wheeze andatopy in childhood. Thorax 2008;63(6):507–13.

[93] Heederik D, von Mutius E. Does diversity of environmental microbial exposurematter for the occurrence of allergy and asthma? J Alergy Clin Immunol2012;130(1):44–50.